Advances in Physical Processing of Cathode and Anode Materials from Spent Lithium-Ion Batteries

Abstract

The rapid expansion of lithium-ion battery (LIB) applications and the imminent surge in end-of-life batteries have intensified the demand for efficient, scalable recycling technologies. Physical separation of cathode and anode materials is a crucial pretreatment step that enables high-value metal recovery and direct material regeneration. This review critically examines recent advances in three major physical separation technologies—magnetic separation, gravity separation, and flotation—for processing spent LIB electrodes. Rather than offering a descriptive summary, the review systematically analyzes separation mechanisms, key controlling parameters, and pretreatment strategies across representative cathode chemistries, including LiFePO₄ (LFP), LiCoO₂ (LCO), and Ni–Co–Mn (NCM) systems. Particular emphasis is placed on emerging flotation-enhancement strategies, such as nanobubble-assisted and ultrasonic-enhanced flotation, and their underlying mechanistic roles in improving selectivity and recovery. Comparative evaluation indicates that magnetic separation has reached industrial maturity for LFP–graphite systems but remains chemistry-specific. Gravity separation is effective for coarse particles and centrifugal-assisted graphite recovery yet shows limited selectivity for fine particles. Flotation has become the dominant research focus for complex, fine-particle separations due to its tunable surface chemistry. Despite significant laboratory progress, challenges remain, including incomplete binder removal, limited understanding of electrode surface reconstruction during pretreatment, fine-particle entrainment, and the gap between bench-scale research and industrial implementation. Future research priorities include green reagent development, intelligent separation control, and integration with direct regeneration routes to advance closed-loop LIB recycling towards sustainable development.

1. Introduction

Lithium-ion batteries (LIBs) have become core energy storage devices in the new energy industry due to their excellent properties, such as high energy density, long cycle life, and low self-discharge rate, being widely used in new energy vehicles, smartphones, energy storage stations, and other fields [1]. In recent years, with the rapid development of the global new energy vehicle industry and the first batch of LIBs entering the retirement period, the output of spent LIBs has shown explosive growth [2,3,4]. According to statistics, the global expected retirement volume of spent LIBs will exceed 2 million tons by 2025, and the global LIB energy storage capacity will exceed 2500 GW·h (approximately 12.7 million tons) by 2030 [5]. Spent cathode materials from LIBs contain valuable strategic metal resources such as lithium (Li), cobalt (Co), and nickel (Ni) (Li: 5–7 wt.%, Co: 5–20 wt.%, Ni: 5–10 wt.%), which are far higher than the corresponding contents in natural ores (e.g., Co content in LiCoO₂ is about 850 times that in typical associated cobalt ores) [6]. Meanwhile, spent LIBs contain toxic and harmful substances such as electrolytes and binders [7]. Improper disposal not only causes the waste of strategic resources but also threatens the ecological environment and human health through soil and water pollution [5,8,9]. Therefore, achieving efficient recycling of spent LIBs is not only an important way to alleviate the shortage of strategic resources and ensure supply chain security but also an inevitable requirement for practicing the concept of green development.

Currently, the main technical routes for the resource recovery of valuable metals from spent LIBs include pyrometallurgy, hydrometallurgy, and the emerging direct regeneration technology [10,11]. Pyrometallurgy extracts alloys through high-temperature smelting, but has high energy consumption and difficulty in recovering Li [7]. Hydrometallurgy based on acid/alkali leaching, separation, and purification is relatively mature, with a high recovery rate, but it usually has a long process, high reagent consumption, and heavy pressure on wastewater treatment [8]. Direct regeneration technology aims to repair the structure of failed cathode materials, enabling higher value-added recovery, but it has extremely strict requirements on raw material purity [5]. The optimization of pretreatment processes has a crucial impact on the efficiency of subsequent hydrometallurgical or pyrometallurgical recovery [12]. However, these technical routes all face a common pretreatment bottleneck when dealing with complex battery black mass (referring to the mixture of electrode materials obtained after crushing and dismantling): the efficient separation of cathode materials (e.g., LiCoO₂ (LCO), Ni-Co-Mn (NCM), and LiFePO₄ (LFP), etc.) and anode materials (mainly graphite). If the two are not effectively separated, graphite will consume a large amount of acid and produce harmful gases during hydrometallurgical leaching, increasing reagent costs and environmental risks; in the pyrometallurgical process, it acts as a reducing agent that is difficult to control accurately, affecting metal yield and quality; for direct regeneration processes, the presence of graphite is considered an unacceptable impurity because residual graphite can interfere with lithium replenishment reactions, hinder crystal structure repair, and introduce carbon-related side reactions during high-temperature treatment, ultimately deteriorating electrochemical performance and product consistency [13]. In addition, graphite contamination may reduce tap density and compromise the structural integrity of regenerated cathode materials. Therefore, developing efficient, environmentally friendly, and low-cost cathode–anode material separation technologies is a key pretreatment step to improve the efficiency of subsequent valuable component extraction, reduce overall recovery costs, and achieve high-quality direct regeneration. Its technical level directly determines the economy and sustainability of the spent LIB recycling industry.

In this context, physical separation technologies have attracted much attention due to their potential for low cost, environmental friendliness, and easy scaling-up. This review focuses on three mainstream physical separation technologies—magnetic separation, gravity separation, and flotation, which can avoid or reduce the use of chemical reagents and the generation of secondary pollution. Although emerging physical separation methods such as electrostatic separation and air classification have attracted increasing research interest, their industrial applicability remains limited due to challenges, including feed size sensitivity, moisture dependence, and insufficient large-scale validation. Therefore, this review primarily focuses on magnetic separation, gravity separation, and flotation, which currently demonstrate higher technological maturity and industrial relevance. Among them, flotation has become a current research hotspot due to its efficient treatment capacity and separation selectivity for fine particles. Several thematic reviews have conducted in-depth discussions on its interfacial mechanisms, reagent systems, and pretreatment technologies [11,13,14]. Ji et al. [15] pointed out in their review that a single physical separation process is difficult to cope with complex electrode systems, and the problem of surface property homogenization caused by binder residues has not been fully solved, seriously affecting separation selectivity. Efendi et al. [16] further emphasized that existing studies have not thoroughly explored the microscopic mechanisms such as the surface charge (zeta potential) and functional group evolution of cathode and anode materials, leading to a lack of theoretical support for the selective regulation of flotation reagents, and the differences in flotation behavior of different cathode systems (LFP/NCM/LCO) have not been fully utilized. Magnetic separation and gravity separation show unique advantages in specific cathode systems (e.g., LFP) or large-scale roughing scenarios. Systematically reviewing and evaluating research progress, mechanisms, technical advantages and disadvantages, and applicability limits of these physical separation methods is of great significance for promoting the development of spent LIB recycling technologies towards refinement, environmental sustainability, and higher value addition. Furthermore, the refined optimization of magnetic separation and gravity separation technologies (e.g., adaptation of high-gradient magnetic separation parameters, construction of heavy liquid recovery systems) also needs further advancement.

This review was prepared in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The study design, the literature search, screening process, and data synthesis followed the PRISMA framework to ensure transparency, rigor, and reproducibility (Supplementary File S1). The identification of studies via databases and registers using the PRISMA 2020 flow diagram is displayed in Figure 1.

In this review, terms such as “industrial maturity” and “scalability” are used in a relative sense, referring to reported pilot-scale demonstrations, throughput ranges, and process integration feasibility rather than fully disclosed commercial CAPEX or OPEX values. Available studies indicate that magnetic separation is already applied at ton-per-hour scales, whereas centrifugal gravity separation and flotation are mostly reported at laboratory or semi-pilot scales (<100 kg/h). Due to limited public disclosure, energy consumption and cost metrics are discussed qualitatively rather than quantitatively.

2. Magnetic Separation

Magnetic separation achieves cathode–anode separation based on the difference in magnetic properties of materials. The anode graphite is a non-magnetic substance, while some cathode materials (e.g., LFP) have natural weak magnetism and can be directly separated by a magnetic field; non-magnetic cathode materials such as NCM and LCO need to undergo prereduction roasting or carbothermal reduction treatment to convert transition metals such as Ni and Co into ferromagnetic metal simple substances or alloys, forming a significant magnetic difference with non-magnetic graphite, thereby realizing the separation [15].

2.1. Key Influencing Parameters

2.1.1. Magnetic Field Characteristics

LFP separation requires a magnetic field intensity of 0.1–1.0 T, and high-gradient magnetic fields can improve the capture capacity of weakly magnetic particles. When the magnetic field intensity increases from 0.2 T to 1.0 T, the grade of graphite concentrate increases from 43 to 91%, but the recovery rate decreases from 86.65 to 69.70% [15,17].

2.1.2. Equipment Operation Parameters for Magnetic Separation

For high-gradient magnetic separation, a pulsation stroke of 12 mm and a frequency of 200 rpm can reduce particle agglomeration and improve separation efficiency. The rotational speed of the drum magnetic separator needs to match the material flow rate to ensure sufficient adsorption of magnetic particles. In industrial applications, the Akkuser process achieves an Fe recovery rate of over 95% through two-stage magnetic separation [18].

2.1.3. Material State Parameters

Cutting the electrode sheet into 1 cm² small pieces before ball milling can avoid the embedding of LFP fine particles into graphite, significantly improving the concentrate grade [17]. Ultrasonic dispersion (200 W, 20 min) can break the agglomeration of magnetic particles. For the mixed system of 500–100 nm Co particles and graphite, the Co recovery rate can reach 99.01–100% [15,19].

2.2. Pretreatment Technologies

The universality and separation performance of magnetic separation are highly dependent on pretreatment processes. A schematic of pretreatment processes is displayed in Figure 2. The optimization directions for different cathode material systems are as follows:

(1)

LFP System

No magnetic conversion pretreatment is required; only crushing, ball milling, and ultrasonic dispersion are needed. The key is to avoid the embedding of fine particles into graphite, which can be achieved by stepwise cutting-ball milling treatment [17].

(2)

NCM/LCO System

Carbothermal reduction or prereduction roasting (600–800 °C, inert atmosphere) is required to convert Ni and Co into magnetic simple substances or alloys, avoiding magnetic failure caused by the oxidation of transition metals [20]. Wu et al. [21] adopted carbothermal reduction under nitrogen atmosphere and wet magnetic separation, achieving Ni and Co recovery rates of over 94% [15,22].

(3)

Fine Particle System

Here, fine particles refer to electrode materials with particle sizes below approximately 100 μm, for which magnetic capture efficiency decreases due to reduced magnetic force and increased surface effects. An “ultrasonic dispersion–magnetic separation” synergistic process is adopted. Ultrasonic power of 200–300 W and dispersion time of 20–30 min can effectively break particle agglomeration and improve the recovery rate of fine magnetic particles [19]. This technology has been extensively studied in the cathode–anode separation of LFP and graphite, forming a variety of targeted processes: high-gradient magnetic separation at a magnetic field intensity of 0.8 T achieves a graphite concentrate grade of 98.6% and recovery rate of 80.01%, and an LFP concentrate grade of 93.3% and recovery rate of 98.69%; wet magnetic separation can effectively separate LCO, NCM from graphite, with a Co recovery rate of 95.72% and a graphite recovery rate of 91.05% [23,24]. In industrial applications, the Akkuser process combines mechanical crushing and magnetic separation, with an annual processing capacity of 4000 tons, no hydrometallurgical or pyrometallurgical steps, and a simple and environmentally friendly process [18]. In addition, related equipment parameter optimization and magnetic field simulation studies have also provided theoretical support for the improvement of magnetic separation processes [25].

Its advantages are excellent LFP–graphite separation performance (recovery rate/grade over 98%), high degree of automation, low environmental pollution from dry (or partial wet) separation, and suitability for industrial large-scale recycling. Its limitations are that NCM, LCO, and other systems require pre-reduction roasting, increasing energy consumption and process complexity. Fine particle agglomeration affects separation efficiency, requiring auxiliary means such as ultrasonic dispersion; high-gradient magnetic separation equipment has high investment costs, and the separation performance for cathode systems such as LiMn₂O₄ (LMO) still needs optimization [15]. From an industrial perspective, magnetic separation is associated with relatively low energy consumption and minimal reagent requirements, which supports its scalability. However, its effectiveness remains highly dependent on sufficient particle liberation and magnetic susceptibility differences.

3. Gravity Separation

3.1. Separation Principle

Gravity separation realizes separation based on the density difference between cathode and anode materials, as illustrated in Figure 3. Graphite has a density of approximately 2.2 g/cm³, while cathode materials such as NCM, LFP, and LCO have a density of 3.6–5.1 g/cm³, with a significant density difference between the two [26]. In the medium of air flow, water flow, or centrifugal force field, particles with different densities are subjected to different forces such as gravity, medium buoyancy, resistance, or centrifugal force. Graphite with lower density is prone to suspension or staying in the upper layer, while cathode materials with higher density are prone to sedimentation or staying in the lower layer, thereby achieving cathode–anode separation [5].

From the perspective of particle settling velocity, the introduction of centrifugal force fields and heavy liquids is crucial to overcoming the limitations of conventional gravity separation. According to Stokes’ law, the settling velocity of fine particles (typically 5–20 μm for electrode materials) in a gravitational field is extremely low due to the constraints of gravitational acceleration, leading to prolonged separation time and poor efficiency [27]. The centrifugal force field significantly enhances the effective acceleration acting on particles, which is directly proportional to the settling velocity, thereby drastically shortening the separation time from hours to minutes or even seconds [28,29,30]. Meanwhile, heavy liquids, such as sodium polytungstate (SPT, Grub am Forst, Northern Bavaria, Germany), Clerici solution (Rome, Italy), with adjustable density (1.0–5.0 g/cm³) not only provide a density gradient that matches the density difference between cathode and anode materials but also optimize the viscosity of the separation medium. This not only ensures that graphite (lower density) floats stably and cathode materials (higher density) settle efficiently, but also avoids excessive settling velocity differences caused by particle size variations, improving separation selectivity for fine and ultrafine particles [31,32].

3.2. Separation Equipment and Key Parameters

3.2.1. Core Equipment

Core separation equipment includes traditional dry equipment (Industrial air flow fabric shaker, Wuxi, China; Metso air classifier, York, PA, USA; Zigzag air flow separator, Los Angeles, CA, USA) and enhanced gravity separation equipment (Falcon centrifugal separator, Vancouver, BC, Canada; Knelson separator, Langley, BC, Canada) [33,34]. Among them, the air flow fabric shaker is suitable for conventional gravity separation of coarse particles, and the Falcon centrifugal separator is adapted for cathode–anode separation of fine particles due to the enhanced centrifugal force field; heavy liquid centrifugal separation commonly uses SPT and Clerici solution as separation media [31,35].

3.2.2. Medium Parameters

The optimized range of air flow velocity is 1–5 m/s; in the centrifugal force field, the centrifugal acceleration is 50–300× g, the density of SPT heavy liquid can be adjusted by dilution or evaporation (1.0–3.1 g/cm³ at 25 °C), and the ternary system (SPT/NaCl/water) can form a steeper density gradient with lower viscosity than traditional heavy liquids (e.g., cesium chloride) [35]; the density of the Clerici solution can be adjusted to 4360–5000 kg/m³ (90 °C). The separation efficiency in the centrifugal force field is closely related to the hydrodynamic behavior [36].

3.2.3. Equipment Operation Parameters for Gravity Separation

The vibration frequency of the air flow shaker is 10–50 Hz, and the bed inclination angle is 5–20°; the rotation speed of the Falcon separator is 4800 rpm (corresponding to centrifugal force of 50–300× g), the fluid water pressure is 0.01–0.05 MPa, and the separation performance is optimal when the feed mass is 300–600 g [28,37]; the separation time can be shortened to 0.2 s when the heavy liquid centrifugal rotation speed is 10,000 rpm [31].

3.2.4. Material Characterization Parameters

The material needs to have a uniform particle size and sufficient dissociation to avoid “graphite–cathode material” composite particles; the separation efficiency is the highest when the particle size is 0.045–0.09 mm, and the separation performance decreases significantly when the particle size is less than 45 μm [32], which was illustrated in Figure 4a.

3.3. Gravity Separation Progress

3.3.1. Air Flow Separator

Research on the application of traditional gravity separation technology in cathode–anode material separation mainly focuses on coarse particle systems. The air flow shaker can achieve significant separation performance under optimized parameters. For example, for NCM–graphite mixed powder, when the air flow velocity is 2.5 m/s, and the vibration frequency is 30 Hz, the cathode material grade is 92.3%, the recovery rate is 91.5%, the graphite grade is 93.1%, and the recovery rate is 92.7% [38,39]. The Zigzag air flow separator can efficiently separate current collectors and diaphragms with a recovery rate close to 100%, but it is not suitable for the separation of fine particle electrode materials [40].

3.3.2. Falcon Centrifugal Separation

Zhan et al. [28] used a Falcon UF separator with water as the medium. When sorting the NCM–graphite mixed system in a single stage, the NCM recovery rate is 93.78%, and the grade is 86.48%; after multi-stage separation (roughing—two-stage cleaning—two-stage scavenging), the NCM purity reaches 99%, and the graphite purity reaches 99.65%; when processing the “black mass” of spent batteries, the cathode material grade exceeds 98% after three separations, with no residual graphite particles. Zhu et al. [32] found that increasing centrifugal force reduces the overflow yield but improves graphite purity, while increasing fluid water pressure has the opposite effect. Moreover, smaller particle sizes result in lower separation efficiency, and the LiCoO₂ grade and recovery rate are optimal for particles of 45–90 μm. Zhang et al. [37] determined the optimal process parameters as water back pressure of 0.025 MPa and rotation speed of 50 Hz, at which the LiCoO₂ grade is 84.87%, and the recovery rate is 83.14%, significantly better than traditional gravity separation.

3.3.3. Centrifugal Separation with Heavy Liquid

While centrifugal gravity separation is often described as environmentally benign due to its physical nature, the environmental footprint of heavy media, including toxicity, recyclability, and loss rates, remains insufficiently quantified.

(a)

SPT Heavy Liquid System

Plewinsky et al. [35] confirmed that SPT can form binary (SPT/water) and ternary (SPT/NaCl/water) systems. The saturated solution density at 25 °C reaches 3.12 g/cm³, the pH stable range is 2–10, the chemical properties are stable with no obvious ion balance change, the viscosity is lower than that of traditional heavy liquids such as cesium chloride, and the cost is more advantageous, suitable for density gradient centrifugal separation of minerals and powder materials. Its ternary system (SPT/NaCl/water) can eliminate sedimentation potential by adding sodium chloride, form a steeper density gradient, and improve separation efficiency.

Al-Shammari et al. [31] used SPT heavy liquid (density 2.4 g/cm³) combined with centrifugal force (10,000 rpm) to achieve rapid separation of graphite from LFP, LMO, NCA, NCM, and LCO, with a separation time of only 0.2 s, graphite purity exceeding 99%, cathode material purity also exceeding 99%, and cross-contamination rate <1%. A schematic was displayed in Figure 4b. The separated graphite was regenerated through NMP cleaning (removing Polyvinylidene Fluoride (PVDF) binder), boric acid treatment, and sintering at 950 °C under nitrogen atmosphere, and its electrochemical performance was significantly restored, with a specific capacity of 340 mAh/g, close to commercial graphite (342 mAh/g), and excellent cycle stability. This study confirms that gravity separation can not only achieve efficient separation but also provide high-purity raw materials for direct graphite regeneration, improving the level of resource high-value-added utilization.

(b)

Clerici Solution System

Al-Shammari et al. [27] used Clerici solution as the heavy liquid, adjusting the density (3800–4800 kg/m³) through temperature to sequentially separate LFP, LMO, LiNiₓCoᵧAl₂O₂ (NCA), NCM, and LCO. The single-step separation time is <10 s, the product purity is high without damaging the material structure, but attention should be paid to its toxicity and recovery cycle.

Traditional gravity separation is dry, environmentally friendly, large in processing capacity, and low in cost. Centrifugal gravity separation (Falcon separation, heavy liquid centrifugation) solves the problem of fine particle separation, achieving high processing accuracy (grade/recovery rate over 98%). The SPT heavy liquid is green, and the separated graphite can be directly regenerated (340 mAh/g), meeting the demand for resource high-value-added utilization [10]. Nevertheless, systems involving dense or heavy media require careful management of medium recovery, loss, and potential environmental footprint during large-scale operation, which should be considered when evaluating their industrial practicality. Traditional gravity separation has poor fine particle separation performance (<10 μm). Centrifugal gravity separation equipment is costly. Clerici solution poses toxic risks and requires supporting recovery systems. SPT heavy liquid shows increased viscosity at high concentrations, which may affect the separation rate. Moreover, it has strict requirements on the material dissociation degree and particle size, and composite particles will seriously deteriorate the separation performance.

3.3.4. Comparison of Different Separators

Compared with traditional air-flow separators, centrifugal separators introduce enhanced acceleration fields that significantly improve the settling velocity of fine particles. Structurally, air-flow systems rely primarily on aerodynamic classification, whereas centrifugal devices integrate rotating bowls and fluid back-pressure to achieve density-based stratification under high G-forces. This fundamental difference explains their distinct applicability to coarse versus fine particle systems.

4. Flotation Separation

4.1. Separation Principle and Difficulties

Flotation is a core technology for separating cathode and anode materials based on the difference in surface hydrophobicity. Anode graphite is naturally hydrophobic, while cathode materials (LFP, LCO, NCM, etc.) are inherently hydrophilic, and efficient separation can theoretically be achieved through bubble adsorption. Its core principle follows Young’s equation. After pretreatment and debinding, graphite exposes hydrophobic basal planes, and cathode materials exhibit hydrophilic properties: after adding reagents to the pulp, collectors enhance the hydrophobicity of graphite, making it selectively attach to bubbles. Depressants inhibit the floating of cathode materials, and frothers stabilize bubbles. Finally, graphite floats up with the foam to form concentrate, and cathode materials remain in the pulp to become tailings, realizing non-destructive separation of cathode and anode [11,16]. The principle of froth flotation was displayed in Figure 5.

Its practical application faces three core difficulties [11,16,41]: (1) organic binders such as PVDF and carboxymethyl cellulose (CMC) cover the electrode surface, homogenizing hydrophilicity and hydrophobicity and reducing the surface difference between cathode and anode materials (when PVDF exists, the contact angle difference between the two is only 6.7°) [42]; (2) the solid electrolyte interphase (SEI) film formed during battery cycling (containing LiF, ROCO₂Li, etc.) will reduce the hydrophobicity of graphite (contact angle decreases from 65.4 to 40.5°) [43]; and (3) electrode particles <10 μm have a large specific surface area and high surface energy, making them prone to agglomeration and leading to decreased particle–bubble attachment efficiency. In addition, the pH dependence of the surface charge characteristics (zeta potential) of cathode and anode materials affects the selectivity of reagent adsorption, and the separation efficiency decreases significantly if not fully regulated [16]. Related surface potential regulation studies provide a theoretical basis for flotation separation [44]. It is worth noting that when flotation is used to separate different types of cathode materials (e.g., LCO and LMO), it also faces the challenge of “heterocoagulation”. Due to the fact that the surface charges (e.g., zeta potential) of different cathode materials may have the same sign and small absolute value under similar pH (e.g., both are weakly negatively charged at pH 7–9), the electrostatic repulsion between particles is insufficient, leading to their mutual agglomeration, which seriously deteriorates the selective flotation separation performance [45]. The differences in crystal structures of different cathode systems (e.g., olivine structure of LFP, layered structure of LCO) will lead to differences in the number of surface hydroxyl groups and active sites, thereby affecting reagent adsorption and flotation behavior. This difference needs to be addressed through targeted process design [13]. In addition, recent studies have revealed that different cathode materials exhibit distinct flotation behaviors due to their inherent physicochemical properties (e.g., particle size, specific surface area, surface functional groups), which increases the complexity of separation in mixed systems. For example, hydrophilic LFP may exhibit abnormally high floatability in mixed flotation systems due to its ultrafine particle size and high specific surface area, even superior to naturally hydrophobic graphite, which poses a new challenge to the separation strategy based on the simple “hydrophobic-hydrophilic” dualism [46].

4.2. Pretreatment to Remove Organic Binder

Binder removal is a prerequisite for efficient flotation, which needs to restore the inherent surface properties of electrode materials. The adaptability of pretreatment technologies for different cathode systems is illustrated as follows.

4.2.1. Roasting (Pyrolysis) Treatment

The LFP system requires roasting at 500 °C for 1 h to completely decompose PVDF, and Fe²⁺ oxidation should be avoided. For the LCO system, the contact angle difference between cathode and anode is the largest (35° vs. 75°) when roasted at 450 °C for 15 min, with the best separation performance. The roasting temperature for the NCM system should be controlled at 400–500 °C to prevent the oxidation of transition metals [11,15,47]. Its advantages are simple operation and high binder removal efficiency. Its limitation is high energy consumption, and harmful gases such as HF need to be absorbed with alkaline solutions.

It is suitable for the LFP system (inert atmosphere 400–600 °C), which can avoid Fe²⁺ oxidation and recover pyrolysis oil and gas (short-chain hydrocarbons) at the same time, with a fluorine residue as low as 0.067 wt.%, and the pyrolysis temperature for NCM and LCO systems needs to be slightly higher than that for LFP (500–550 °C) to ensure complete decomposition of organic binders without damaging the electrode structure [48].

4.2.2. Fenton Oxidation

It is more suitable for the LCO system. When Fe²⁺:H₂O₂ = 1:120 and liquid–solid ratio = 75:1, hydroxyl radicals can efficiently destroy the PVDF structure, degrading the organic film within 30 min. For the LFP system, due to the interference of Fe sites, the pH should be strictly controlled (around 3.0) to avoid secondary pollution [11,49]. Fe(OH)₃ secondary products need to be dissolved by adjusting the pH with hydrochloric acid to avoid affecting flotation.

4.2.3. Mechanical Grinding

For the LCO system, when ground for 5 min, the concentrate grade reaches 97.19%, but the recovery rate is only 49.32%; thus, the grinding intensity should be controlled. For the LFP system, due to the fine particle size (D₅₀ = 1.38 μm), excessive grinding is prone to agglomeration, requiring ultrasonic dispersion. Cryogenic grinding (−196 °C) is suitable for the NCM system, which can enhance the brittleness of binders, increase binder removal efficiency by 30%, and cause no secondary pollution [50,51,52].

4.2.4. Solvent Dissolution

NMP can achieve a PVDF stripping efficiency of 99% within 90 min under ultrasonic assistance at 70 °C and 240 W, suitable for various systems. Deep eutectic solvents (DESs), such as the choline chloride–glycerol system, can decompose PVDF at 190 °C, which is green and environmentally friendly and can be recycled, making it more suitable for LFP and NCM systems with high environmental requirements [16,53].

4.2.5. Plasma Treatment

It is suitable for NCM and LCO systems [54,55]. The optimal parameters are power 150 W, gas flow rate 300 mL/min, and treatment time 3 min. In their work, the anode contact angle increases from 65.70 to 113.33°, and the cathode material contact angle decreases from 40.70 to 26.17°, with the flotation selectivity index increasing from 2.41 to 4.28. For the LFP system, due to the rich surface hydroxyl groups, flotation should be carried out quickly after plasma treatment to avoid surface re-hydrophilization [54,55].

4.3. Flotation Reagent Optimization

4.3.1. Collectors

Kerosene (300 g/t) and n-dodecane (250 g/t) are commonly used reagents. Emulsified kerosene has stronger selectivity for LCO. When the pulp pH is 9, and the concentration is 5%, the LCO recovery rate is 90.12%, and the grade is 88.03%. N-dodecane is suitable for the LFP system. When combined with 150 g/t MIBC and pulp concentration of 7.5%, the cathode material grade is 96.80%, and the recovery rate is 95.26%. The NCM system is suitable for E_SCAID^TM 110, which can enhance graphite hydrophobicity while reducing non-selective adsorption on NCM [13,56]. However, for the LFP–graphite system, since LFP may show a tendency to compete with graphite for floating under the condition of no collector or specific conditions, the selection and dosage of collectors need to be particularly cautious, and they usually need to be combined with efficient depressants to enhance selectivity [46]. In addition, the development of new green collectors is also a research hotspot. For example, a sustainable graphite flotation was developed using natural oil as a froth flotation collector, demonstrating proper selectivity and environmental friendliness [57].

Natural reagents such as tea saponin and plant extracts are low-toxic and easily degradable. They can replace traditional hydrocarbon reagents in the NCM system, with a graphite recovery rate of 89% and a 20% reduction in reagent dosage. Sodium oleate (NaOl) enhances hydrophobicity by adsorbing on the graphite surface, and can increase the graphite recovery rate to 92% in the NCM system [16].

4.3.2. Frothers

Methyl isobutyl carbinol (MIBC) is the most widely used frother reagent. The optimal dosage for the LFP system is 150 g/t, for the LCO system is 90–120 g/t, and for the NCM system is 120–150 g/t. Combined with kerosene, it can increase the cathode material recovery rate of various systems to over 95% [58]. 2-octanol is suitable for NCM and LMO systems, with the highest flotation recovery rate of cathode materials (87%), while MIBC is the worst (56%), and frothers can enhance hydrophobicity by reacting with surface functional groups of cathode materials [59]. Terpene oil has high foaming efficiency. When the dosage is 100 g/t, the graphite recovery rate reaches 98%, which is suitable for the LCO system. However, the foam viscosity is relatively large, and the dosage should be controlled to avoid entraining cathode particles [11,60].

4.3.3. Depressants and Dispersants

(1)

Depressants

Given the abnormal high floatability of LFP in mixed flotation, the development and application of efficient and selective depressants (such as soluble starch, CMC) are particularly critical for its separation from graphite [46]. Soluble starch (optimal dosage 600 g/t for the LFP system) enhances hydrophilicity by selectively adsorbing on Fe sites on the LFP surface, resulting in an LFP concentrate grade of 84.33% and a recovery rate of 91.57%. CMC (carboxymethyl cellulose) is suitable for NCM and LFP systems. After treatment with 100 mg/L 25 mPa·s CMC, the graphite recovery rate is 96.37%, and the LFP recovery rate decreases to 3.47%. The LCO system is suitable for H acid, which improves the separation selectivity of LCO and graphite by π-π interaction adsorbing on the graphite surface [13,61,62].

(2)

Dispersants

Dispersants are crucial in solving the “heterocoagulation” problem between cathode materials [45]. For example, when separating mixed cathode materials of LCO and LMO, since both are weakly negatively charged in the pH range of 7–9, they are prone to heterocoagulation, resulting in extremely poor flotation selectivity. Studies have shown that adding sodium metasilicate as a dispersant can significantly increase the absolute value of the negative zeta potential on the surface of LCO and LMO particles, enhance the electrostatic repulsion between particles, thereby effectively depolymerizing agglomerates and greatly improving separation efficiency [45]. In addition, NaPO₃ (2.5 × 10⁻⁵ mol/L) can effectively disperse LFP particle agglomeration, increasing the concentrate recovery rate to 91.57%. The NCM system is suitable for sodium silicate (100 g/t), which can reduce fine particle entrainment [61].

4.4. Advanced Auxiliary Strengthening Technologies

4.4.1. Ultrasonic-Assisted Technology

Ultrasonic technology has been widely used in the recovery process of cathode and anode materials [63,64,65]. Through cavitation effect and shear action, ultrasonic technology strengthens binder removal and particle dispersion and is often combined with pyrolysis to form a “pyrolysis–ultrasonic-assisted flotation” process. For example, Zhang et al. [37] adopted pyrolysis at 500 °C for 15 min, followed by ultrasonic cleaning to remove residual organic matter, increasing the grade of LCO cathode material from 67.25% to 93.89% and the recovery rate from 74.62% to 96.88%. In the LFP system, ultrasonic dispersion (200 W, 20 min) combined with flotation can achieve an LFP concentrate recovery rate of 98.69%. After ultrasonic treatment of the NCM system, the particle agglomeration rate decreases by 40%, and the bubble–particle attachment efficiency increases by 25% [11]. Considering the high energy consumption limitation of ultrasound, systematic research can be carried out in the future from aspects such as ultrasonic transducer design and dual-frequency ultrasound combination [66,67].

4.4.2. Nanobubble-Assisted Technology

Nazari et al. [47,60] conducted a series of studies on nanobubble-assisted flotation of graphite. The strengthening mechanism of nanobubble-assisted flotation lies in the formation of nanobubbles (50–200 nm) that coat the surfaces of mineral particles and act as effective “bridges” between particles and conventional flotation bubbles, thereby substantially enhancing bubble–particle attachment efficiency. Simultaneously, nanobubbles promote better selective collector adsorption, enabling a reduction in reagent consumption [66,68,69,70,71]. Experimental results demonstrate that nanobubbles increase graphite flotation recovery by approximately 15% and the flotation rate constant by 33%. In the LCO system, a two-stage nanobubble-assisted flotation process improves the concentrate grade from 65 ± 2% to 93 ± 3%. For the separation of NCM black mass under optimized conditions (pulp concentration 40 g/L, kerosene 300 g/t, and MIBC 150 g/t), the graphite concentrate achieves a grade of 89.22% and a recovery of 85.83%, representing increases of 6.92% and 13.96%, respectively, compared with conventional flotation. In the LFP system, where particles are particularly fine, nanobubbles enhance separation efficiency by more than 20% [58]. Notably, a comprehensive understanding of the stability mechanisms and interfacial behaviors of both bulk and interfacial nanobubbles is essential for elucidating the fundamental mechanisms underlying flotation enhancement [66,70,71].

4.4.3. Surface Analysis Technology-Assisted Optimization

Surface analysis techniques—including Fourier transform infrared (FTIR) spectroscopy, X-ray photoelectron spectroscopy (XPS), and time-of-flight secondary ion mass spectrometry (ToF-SIMS)—provide critical theoretical support for optimizing flotation processes across different systems. FTIR enables the identification of PVDF decomposition, as evidenced by the disappearance of the CF₂ vibrational peak at 175 °C. XPS allows quantitative assessment of fluorine content, revealing a 52% reduction following roasting at 400 °C. Meanwhile, ToF-SIMS offers detailed insights into binder distribution, demonstrating that CMC preferentially migrates to the electrode surface compared with styrene–butadiene rubber (SBR) [11,16]. By elucidating binder residue characteristics and reagent adsorption mechanisms, these surface analysis methods enable precise optimization of pretreatment temperatures and reagent dosages tailored to different flotation systems.

4.5. Research Progress on Combined Processes

A single flotation technology is difficult to cope with complex systems, and combined processes have become a research hotspot.

4.5.1. Flotation–Magnetic Separation Combination

For the LFP system, the “magnetic separation roughing-flotation cleaning” process is adopted, with a concentrate grade of 95.2% and a graphite grade of 92.0%, which is 4.7% higher than single magnetic separation. For the NCM system, through “flotation decarburization–magnetic separation enrichment”, the cathode material recovery rate reaches 94.3% [13,15,17,22,72,73].

4.5.2. Pyrolysis–Flotation–Gravity Separation Combination

For the NCM-LFP mixed system, a “low-temperature pyrolysis–flotation separation of graphite–gravity separation enrichment of cathode materials” is adopted, with a total cathode material recovery rate of 94.3% and graphite purity exceeding 99% [15,52].

4.5.3. Coarse Flake Particle Flotation Technology

By utilizing the density difference between cathode and anode flakes in the range of 212–850 μm, separation is achieved by balancing gravity and capillary forces without removing binders. This method is suitable for coarse-particle LCO and NCM systems, with a graphite grade of 99% and a recovery rate of 95% [52].

Flotation is a core technology for cathode–anode separation, with extensive research on systems such as LFP, LCO, and NCM [13,46]. By integrating advanced techniques—including plasma-assisted binder removal and nanobubble or ultrasonic enhancement—the grade and recovery of cathode–anode separation can exceed 95%, with particular effectiveness for fine particles (<100 μm) and the ability to achieve “non-destructive recovery”, preserving the structural integrity of electrode materials [11,16]. Its advantages include high separation selectivity, allowing adaptation to multi-cathode systems through tailored reagents and pretreatment; strong fine particle processing capability, overcoming the limitations of gravity separation; enhanced efficiency and environmental performance through nanobubbles, ultrasound, and plasma technologies; and flexibility in combined processes to treat complex mixed systems [14]. Despite its strong adaptability to fine particles, industrial flotation involves reagent consumption, wastewater management, and higher operational complexity, all of which must be balanced against separation performance when considering large-scale deployment. Nevertheless, challenges remain: conventional flotation generates wastewater requiring treatment; some reagents pose environmental risks while greener alternatives are costly; binder removal pretreatments increase process complexity and costs; scaling from laboratory to industrial conditions requires optimization of pulp concentration, reagent dosage, and equipment parameters; and separation processes for certain cathode systems, such as LMO, still need improvement [15]. Flotation generates wastewater streams and reagent residues, but its environmental impact strongly depends on reagent selection and water recycling efficiency.

5. Comparative Analysis and Optimization Discussion

Although both flotation and centrifugal gravity separation are proposed for fine particle systems, their competitive relationship depends strongly on feed characteristics. Centrifugal gravity separation, particularly SPT-based approaches, can rapidly achieve ultrahigh purity (>99%) for well-liberated and narrowly distributed particles but require precise density control and heavy media management. In contrast, flotation exhibits higher adaptability to surface heterogeneity and mixed chemistry, making it more suitable for industrially realistic black mass, albeit with increased reagent consumption and wastewater generation.

5.1. Core Performance Comparison

Based on different separation principles—magnetism, density, and surface hydrophobicity—magnetic separation, gravity separation, and flotation exhibit distinct performance characteristics and application boundaries in cathode–anode material recovery, as summarized in

Table 1. Comparison of different physical separation technologies.

Separation Method	Separation Basis	Core Advantages	Key Limitations	Applicable Scenarios	Typical Performance (Cathode–Anode Separation: Grade/Recovery Rate)	Regeneration Potential
Magnetic Separation	Magnetic difference	Excellent LFP–graphite separation performance (over 98%); high degree of automation; low pollution; mature industrial cases (Akkuser process)	NCM/LCO requires pretreatment; high-gradient equipment is expensive; fine particles need dispersion; poor adaptability to LMO system	Cathode–anode separation of LFP batteries; industrial large-scale recycling	LFP concentrate: 93.3%/98.69%; Graphite: 98.6%/80.01%; Industrial-grade Co recovery rate: 95.72%	Needs further purification and regeneration
Gravity Separation	Density difference (including centrifugal force enhancement)	Traditional gravity separation: dry process, environmental friendliness, large processing capacity, low cost; Centrifugal gravity separation: strong fine particle processing capacity, high purity (over 99%), green SPT heavy liquid; Separated graphite can be directly regenerated (340 mAh/g)	Traditional gravity separation: poor fine particle separation performance; Centrifugal gravity separation: high equipment cost; Clerici solution has toxic risks	Traditional gravity separation: large-scale roughing of coarse particle cathode and anode; Centrifugal gravity separation: fine particle cathode–anode fine separation + graphite regeneration	Traditional gravity separation (NCM-graphite): 92.3%/91.5%; Centrifugal gravity separation (NCM-graphite): 99%/93.78%; SPT heavy liquid separation: graphite 99%/99%	Graphite regeneration performance is close to commercial
Flotation Separation	Surface hydrophobicity difference	High selectivity, suitable for multi-cathode systems; strong fine particle processing capacity; non-destructive recovery; advanced technology-assisted efficiency improvement; flexible combined processes	Produces wastewater; requires pretreatment; high cost of green reagents; industrial parameters need optimization; fine particle entrainment problem	Fine particle cathode–anode fine separation; multi-cathode systems; combined processes for complex mixed systems	LFP-graphite: 96.80%/95.26%; NCM-graphite: 95%+/90%+; Nanobubble assistance: graphite 89.22%/85.83%	Can retain material structure, supporting direct regeneration

. The grade and recovery values in the table come from different experimental setups, including single-stage tests and multi-stage separation circuits. Therefore, direct numerical comparisons should be interpreted with caution. The data are mainly intended to reflect relative technical potential rather than strict performance rankings. It should be noted that many reported grade and recovery values (>98–99%) are obtained from laboratory-scale experiments using simplified feedstocks, controlled particle size distributions, and single-cathode systems. In practical industrial black mass, mixed chemistries, binder aging, electrolyte residues, and particle agglomeration introduce significant variability, which may substantially reduce separation efficiency. Therefore, laboratory-scale metrics should be interpreted as indicators of separation potential rather than direct predictors of industrial performance. Magnetic separation shows excellent selectivity in the LFP–graphite system, achieving grade and recovery rates above 98%, and has matured industrial applications such as the Akkuser process. Its dry operation, high automation, and low pollution make it particularly suitable for large-scale LFP battery recycling. Gravity separation includes traditional and enhanced centrifugal methods: the former offers a dry process, high throughput, and low cost, making it ideal for large-scale roughing of coarse particles, while the latter—especially centrifugal separation using green heavy liquids like SPT—effectively addresses fine particle separation, achieving purities above 99% and integrating well with direct graphite regeneration, highlighting its potential for high-value resource utilization. Flotation, relying on surface property-based selectivity, excels in fine separation of multi-cathode systems (LCO, NCM, LFP), particularly for particles <100 μm, and can achieve non-destructive separation through auxiliary techniques such as nanobubbles and ultrasound, enabling direct regeneration of electrode materials; however, it requires strict pretreatment and faces challenges, including wastewater management and high costs of green reagents. Overall, each technology has its focus: magnetic separation is optimal for industrial-scale processing of specific systems such as LFP, gravity separation is advantageous for roughing and graphite recovery, and flotation serves as the core approach for fine, high-selectivity, multi-system separation. Future developments are expected to emphasize refined selection, multi-technology integration, and intelligent process control based on material characteristics—including cathode type, particle size distribution, and regeneration requirements—to balance separation efficiency, environmental performance, and economic viability [74].

5.2. Technology Selection Criteria

For single-technology selection, magnetic separation is preferred for the LFP system due to its low cost and high efficiency. Centrifugal gravity separation using SPT heavy liquids is suited for cathode–anode separation of fine particles (<10 μm) and graphite regeneration. Traditional gravity separation is ideal for large-scale roughing because of its high processing capacity and environmental friendliness. Flotation, often combined with nanobubble or ultrasonic assistance, is preferred for fine separation of multi-cathode systems. Finally, coarse flake particle flotation (212–850 μm) is effective for coarse particles, offering high efficiency without requiring binder removal [75]. For complex or high-value systems, combined approaches are recommended: for mixed systems such as NCM + LFP + fine particles, a sequence of traditional gravity separation roughing, magnetic impurity removal, and centrifugal gravity separation or flotation cleaning is effective. For direct graphite regeneration, SPT heavy liquid centrifugal separation followed by NMP cleaning and boric acid treatment plus sintering enables high-value recovery. For multi-cathode mixed systems, calcium oxide-assisted pyrolysis with green reagent and nanobubble flotation improves selectivity and environmental performance. And for industrial-scale applications, processes such as Akkuser (magnetic + mechanical separation) or battery resources (magnetic + flotation + hydrometallurgy) can be employed to achieve synergistic recovery of multiple components [18].

6. Challenges and Outlook

6.1. Main Challenges

(1)

Insufficient Pretreatment Accuracy

The widespread existence of “composite particles” is caused by binder residues. Emerging binder removal technologies, such as plasma and calcium oxide assistance, have high costs and limited large-scale applications. In addition, secondary products, such as Fe(OH)_3, affect flotation efficiency [11,15].

(2)

Unclear Surface Modification Mechanism

The correlation between the surface charge (zeta potential), functional group evolution of cathode and anode materials, and reagent adsorption has not been fully clarified. Moreover, the differences in flotation behavior of different cathode systems (LFP/NCM/LCO) have not been fully utilized, thereby restricting selective regulation [13,16].

(3)

Complex Influence of Key Parameters

The influence of parameters such as lithium ion concentration and bubble size varies among different cathode systems, and unified optimization standards are lacking. In addition, the fine particle entrainment problem has not been fully solved in industrial scenarios [10].

(4)

Low Level of Scaling-up and Intelligence

There is a parameter gap between laboratory and industrial applications, and intelligent monitoring and regulation means are lacking. Centrifugal gravity separation equipment has high investment, and the heavy liquid recovery system is not yet mature. Flotation wastewater treatment costs are high [41].

(5)

Balance between Environmental Friendliness and Economy

Factors such as flotation wastewater treatment, green reagent costs, and toxicity of centrifugal gravity separation heavy liquids affect process promotion. Technical bottlenecks, such as nanobubble stability and plasma energy consumption, need to be broken through.

(6)

Insufficient Connection with Regeneration

Although some separation technologies can obtain high-purity products, the connection process with subsequent regeneration processes still needs optimization.

6.2. Future Development Alternatives

(1)

Development of Precise Pretreatment Technologies

The optimization of plasma treatment parameters should promote low-energy-consumption technologies for PVDF decomposition. Recyclable green solvents (such as DESs) should be developed to replace traditional NMP and acid–base reagents. In addition, integrated “crushing–classification–debinding” equipment should be developed to improve the material dissociation degree and uniformity. Surface analysis techniques such as FTIR and XPS should be employed to monitor binder residues in real time and accurately control pretreatment parameters for different systems [11,16].

(2)

Deepening Surface Mechanism and Reagent Optimization

Research should be conducted to clarify the correlation between the surface charge (zeta potential), functional groups of cathode and anode materials, and reagent adsorption, and to design targeted depressants/collectors combined with the crystal structure differences in different cathode systems. Efficient green reagents (such as modified plant extracts and bio-based reagents) should be developed to reduce costs and environmental risks. In addition, nanobubble generation and stabilization technologies should be optimized to strengthen their synergistic effect with reagents [13,41].

(3)

Optimization of Key Parameters and Process Adaptation

Systematically study the influence of laws of parameters such as lithium ion concentration, bubble size, and particle size on different cathode systems, and establish a parameter optimization database; develop selective flocculation technologies for fine particle entrainment problems, combined with nanobubble assistance to improve separation efficiency; promote the application of coarse flake particle flotation technology in coarse particle systems to reduce pretreatment costs [14]. Furthermore, the influence of physical fields, such as ultrasound, on the flotation process (including true flotation and entrainment) should be studied to provide a basis for process regulation [76]. At the same time, with the help of online analysis technologies such as focused beam reflectance measurement (FBRM) and particle video microscope (PVM), in-depth exploration of the interaction behavior between coarse and fine particles in the flotation system is of great significance for understanding and solving key problems such as fine particle entrainment and heterocoagulation [53].

(4)

Artificial Intelligence and Greenization Upgrade

Adopt machine learning to establish separation performance prediction models, and realize automatic parameter adjustment combined with online detection technologies (such as LIBs) [77,78]; develop green reagents such as natural plant extracts; develop flotation wastewater recycling technologies, combined with reagent-free auxiliary technologies such as nanobubbles and ultrasound to reduce environmental impact and costs; optimize the design of nanobubble generators and centrifugal gravity separation equipment structures to improve bubble stability and scaling-up application capabilities; develop heavy liquid recovery and harmless treatment technologies, focusing on promoting the SPT green heavy liquid system to reduce the environmental risks of centrifugal gravity separation.

(5)

Development of Integration of Separation and Regeneration

Promote the in-depth integration of gravity separation and graphite regeneration processes, optimize the “separation–cleaning–sintering” integrated process, and reduce regeneration energy consumption; explore the connection technology of direct regeneration of cathode materials after flotation separation to improve the level of resource high value-added utilization; promote combined processes (such as magnetic separation–flotation–gravity separation) at the industrial end, referring to the scaling-up experience of the Akkuser and Battery Resources processes to achieve synergistic recovery of multiple components [10,18]. Research on the integration and optimization of the entire process chain—from current collector stripping to cathode–anode separation and subsequent material regeneration—should be strengthened [6].

7. Conclusions

Efficient separation of cathode and anode materials from spent lithium-ion batteries is a decisive step toward high value-added resource recovery. The three mainstream physical separation technologies—magnetic separation, gravity separation, and flotation—exhibit distinct technical strengths and application niches. Magnetic separation demonstrates high efficiency and industrial maturity for LFP–graphite systems. Conventional gravity separation is well-suited for large-scale roughing operations, while centrifugal gravity separation, including SPT heavy-liquid systems, offers a promising route for fine-particle separation and graphite regeneration. Flotation, especially when coupled with plasma treatment, nanobubble generation, ultrasonic assistance, and coarse flake particle strategies, has emerged as the core technology for selective separation in complex, multi-cathode systems, with significant potential for process integration.

Despite notable progress, current cathode–anode separation technologies remain constrained by limited pretreatment precision, insufficient mechanistic understanding of electrode surface reconstruction, strong coupling among key operating parameters, and poor adaptability to scale-up and industrial integration. Future research should prioritize precision-oriented pretreatment, in-depth elucidation of surface and interfacial mechanisms, and systematic optimization of critical process parameters. Equally important are the intelligent and green upgrading of separation processes and the consolidation of dispersed technical achievements and operational windows reported in the literature. Breakthroughs are particularly needed in addressing nanobubble stability, reducing plasma energy consumption, and improving heavy-liquid recovery in centrifugal gravity separation. Ultimately, strengthening the coupling between separation and regeneration processes will be essential to achieve economically viable, environmentally benign, and high-efficiency utilization of electrode materials, thereby providing core technical support for the closed-loop recycling industry of spent lithium-ion batteries towards sustainable development.