Lithium-ion batteries (LIBs) are widely applied in electric vehicles and portable electronics due to their high energy density and environmentally friendly characteristics. However, large numbers of LIBs would be retired in the future because of their limited lifespan (5-8 a). Used batteries contain numerous organic materials and valuable elements. They cause serious pollution and resource waste in the absence of suitable treatment. Therefore, typical pre-treatments are carried out to obtain used cathode or anode materials. A series of recycling methods are explored to utilize spent electrode materials, such as element-extracting leaching and direct regeneration. Compared to element-extracting leaching, direct regeneration displays lower energy consumption and higher element recovery, making it regarded as a next-generation approach. In the pre-treatment process, however, the obtained electrode materials are mixed (e.g., LiFePO₄ and graphite), resulting in low leaching efficiency and inferior regeneration performance. Also, direct regeneration processes still suffer from long treatment durations and high energy requirements. It is thus important for recycling used batteries to explore highly effective separation and regeneration methods.

Based on the particle size of raw material, three graded products involved coarse fraction (i.e., +0.21 mm), medium fraction (i.e., -0.21 mm - +0.15 mm), and fine fraction (i.e., -0.15 mm). Magnetic separation experiments on each fraction were conducted, separately. The coarse fraction was processed at a pulsation frequency of 200 times/min and a background magnetic field intensity of 1.0 T, the medium fraction at 240 times/min and 1.2 T, and the fine fraction at 260 times/min and 1.6 T. 5 g of pre-heating waste graphite was placed in 20 mL of hydrochloric acid solution as a stripping solution. After stirring for 5 h, the mixture was washed for three times with deionized water and then further cleaned for three times with an ethanol solution, enabling the short-process regeneration of graphite material.

The pre-classification, magnetic separation, and "surface peeling" regeneration are applied to transform mixed electrode materials into regenerated products. Used LiFePO₄ typically exhibits large bulk sizes, while used graphite particles are comparatively smaller. At the particle size of < 0.075 mm, the carbon content reaches 86.35%. Evidently, simple sieving effectively separates the mixed components. The resulting tailings primarily consist of carbonaceous material (96.53%) and impurities after magnetic separation. To address the detrimental effects of residual impurities, a "surface peeling" regeneration method is developed for used graphite. Increasing the number of treatment cycles progressively tailors surface/near-surface properties, while removing impurities. The related analysis reveals that high-temperature active atmospheres promote an impurity evolution, while acid solutions facilitate their removal. Based on the water-gas shift reaction principles, active carbon atoms react with water, eliminating degraded surface/near-surface layers to yield regenerated graphite. The physicochemical characterization indicates that the regenerated samples exhibit high graphitization, smooth surfaces, and reduced particle sizes. As LIB anodes, these materials deliver a capacity of 320 mA·h·g^-1 at 1.0 C after 300 cycles, with an initial Coulombic efficiency of >80%. At 2.0 C, a capacity retention remains 280 mA·h·g^-1 for over 300 cycles. Increased "surface peeling" cycles shortens activation time due to the optimized surface/near-surface properties, enhancing lithium-ion diffusion. The kinetic analysis indicates that excessive "surface peeling" treatments damage surface structures, while increasing pseudocapacitive contributions. At 0.9 mV·s^-1, pseudocapacitive contributions reach 75%, with a lithium-ion diffusion coefficient of 4.0 × 10^-11 cm²·s^-1, demonstrating a significant physicochemical evolution.

For the promising potential of used LIB graphite, effective recovery methods (i.e., from separation of mixed materials to graphite regeneration) were used. Used LIB materials consisted of large LiFePO₄ bulks and small graphite particles. For pre-classification, the mixed materials were segregated by particle size ranges, i.e., larger fractions were predominantly LiFePO₄, while smaller fractions were primarily graphite. At particle sizes of < 0.075 mm, carbon content reached 86.35%, with LiFePO₄ (12.03%) and some impurities. The magnetic separation increased carbon content to >96%, achieving an effective separation of mixed electrode materials. Subsequently, used graphite was regenerated by "surface peeling" strategie, removing impurities (i.e., conductive carbon, Al, Cu, and LiFePO₄) and tailoring surface/near-surface properties. The optimized sample with a reduced particle size exhibited smooth surfaces and high graphitization. As LIB anodes, these materials delivered the electrochemical performance up to 323 mA·h·g^-1 at 1.0 C. The analysis revealed that tailored diffusion channels played a key role in enhancing reversible lithium-ion storage and deep insertion behaviors. This work could provide effective all-process recycling strategies from pre-classification and separation to regeneration, while elucidating the regeneration mechanism of used graphite.