Separation of Graphites and Cathode Materials from Spent Lithium-Ion Batteries Using Roasting–Froth Flotation

Abstract

The separation of graphites and cathode materials from spent lithium-ion batteries (LIBs) is essential to close the loop of material used in LIBs. In this study, the roasting characteristics of the spent LIB materials are carefully analyzed, and the effects of roasting on the surface morphology and elemental chemical states of electrode materials are fully investigated by thermogravimetric analysis, SEM-EDS, and XPS to explore the roasting–flotation enhancement mechanism. Then, froth flotation is utilized to separate the graphites and cathode materials from the spent LIB materials. The optimal roasting temperature is determined by thermogravimetric analysis and the SEM-EDS analysis of the spent LIB materials. The results suggest that the organic binder can be effectively removed from the spent LIB materials at the roasting temperature of 500 °C, and there is almost no loss of graphite. The XPS results indicate that, in the process of roasting, the decomposition products of the organic binder can easily react with valuable metals (Ni, Co, and Mn) to produce corresponding metal fluoride. The flotation results of the spent LIB materials after roasting at the optimal conditions indicate that graphites and cathode materials can be efficiently recovered through roasting–froth flotation. When the dosage of kerosene is 200 g/t and the dosage of methyl isobutyl carbinol (MIBC) is 150 g/t, the cathode materials grade is 91.6% with a recovery of 92.6%, while the graphite grade is 84.6% with a recovery of 82.7%. The roasting–froth flotation method lays the foundation for the subsequent metallurgical process.

1. Introduction

The number of lithium-ion batteries (LIBs) has increased rapidly because of the vigorous promotion of new energy products [1]. The number of spent LIBs also continues to rise rapidly because more LIBs are being used in large quantities [2]. It is predicted that the total global LIB market size will reach over 97.70 GWh, about 1.52 million tons [3], while the number of spent LIBs is expected to increase to 700,000 tons by 2025 [4]. Spent LIBs contain a large number of valuable metals, including Ni, Co, Mn, and Li, which are exceptionally valuable [5]. Recycling spent LIBs is an essential component to developing sustainable energy storage solutions to combat climate change [6,7]. According to statistics, spent LIBs usually contain 5–20% Co, 5–10% Ni, 5–7% Li, 5–10% other metals (Cu, Al, Fe, etc.), 15% organics, and 7% plastics [8]. Effectively recycling these metals will greatly alleviate the shortage of raw materials for LIBs in the future [9]. Simultaneously, spent LIBs also contain electrolytes, binders, and other harmful substances [10,11]. These will cause great harm to the environment if not handled properly. Thus, the effective recovery of spent LIBs is a key problem for which a solution is urgently required [12].

The main purpose of recycling spent LIBs is to recover electrode materials (cathode materials and graphite) [13]. Currently, the related recycling technologies mainly include pyrometallurgical and hydrometallurgical processes [14,15]. The high-temperature smelting reduction technique is a typical pyrometallurgical process for recycling valuable metals from spent LIBs. The valuable metals are reduced and then recovered in the form of alloys [16]. Spent LIBs are put directly into a smelting furnace without pretreatment. The graphites and organics in the spent LIBs provide heat during combustion, while the metals are reduced and converted to alloys. The obtained alloys are further purified via sulfuric acid (H₂SO₄) leaching and solvent extraction to obtain cobalt oxides (CoO) and nickel hydroxide (Ni(OH)₂). Unfortunately, the graphites and Li elements are either decomposed or wasted [17]. In addition, recycling metals from spent LIBs using the hydrometallurgical method involves inorganic [18] and organic acids [19]. The common inorganic acids used as leaching agents to leach metals from spent LIBs comprise HCl [20], H₂SO₄ [21], HNO₃ [22], and H₃PO₄ [23]. Organic acids include ascorbic acid [24], citric acid [25], oxalic acid [26], acetic acid [27], and so forth. Lee [22] treated LiCoO₂ with 1 mol/L HNO₃ at 75 °C. It was revealed that the leaching efficiencies of Li and Co were only 75% and 40%, respectively. While the leaching rates of Co and Li exceeded 99% when the content of H₂O₂ was 1.7% (v/v), this was mainly because insoluble Co³⁺ was reduced into soluble Co²⁺ in the presence of H₂O₂. Inorganic acids are useful for achieving high metal-leaching efficiency. However, inorganic acids will produce acidic wastewater, Cl₂, SO₂, and other harmful gases, which will cause environmental pollution [19]. Thus, the leaching of metal from spent LIBs with organic acids is environmentally friendly. Chen [28] developed an economically effective approach for the recycling of valuable metals from spent LIBs that combined reduction leaching with selective precipitation. Citric acid was selected as the leaching agent, and D-glucose was used as the reducing agent to dissolve the spent cathode material. The study was conducted under the following conditions: using a 1.5 mol/L citric acid concentration, 20 g/L S/L ratios, 0.5 g/g reducing agent content, an 80 °C temperature, and a 120 min reaction time; consequently, the leaching efficiencies determined with respect to Li, Ni, Co, and Mn were 99%, 91%, 92%, and 94%, respectively. However, the hydrometallurgical process may damage graphite during the leaching process; thus, it may not be suitable for the upcycling of spent anode materials [7].

Before cathode materials are recycled, it is necessary to separate the graphites from spent LIB materials. It can be seen from the literature review that the degrees of hydrophobicity of the cathode materials and the graphites of spent LIB materials are greatly different; for instance, the cathode material shows hydrophilicity, while graphite shows strong hydrophobicity [13,14,29]. Therefore, froth flotation can be used to separate the graphites and cathode materials from spent LIB materials. In the manufacturing process of LIBs, the electrode materials are attached to the current collectors by an organic binder, which leads to difficulty in terms of dissociation between the electrode material and the collectors [30]. As a result, the surface characteristics of electrode materials change, and so it is difficult to achieve the efficient separation of graphites and cathode electrode materials by flotation. At present, researchers have studied the removal of organic binders on the surface of electrode materials and the consequent use of froth flotation to recover graphites and cathode materials. Zhang [31] used pyrolysis to remove the organics on the surface of electrode materials at 500 °C with a heating rate of 10 °C/min and a pyrolysis time of 15 min. Then, the graphites and cathode materials were recovered by froth flotation. Reportedly, with a rougher flotation stage, the underflow product obtained had a grade of 94.7% with respect to cathode materials with a recovery of only 83.7%. However, this method greatly increased the cost due to the implementation of pyrolysis under an inert atmosphere. To improve the separation efficiency of the electrode materials of graphites and cathode materials, the removal of organics is necessary. The dissolution method, the low-temperature roasting method, and pyrolysis are the commonly used organic binder removal methods [32,33]. A solvent dissolution method [34] using an organic solvent weakens the adhesion of the binder of the cathode scraps and removes the binder from spent LIB materials. The organic solvent N-methyl pyrrolidone (NMP) is usually chosen to dissolve the organic binder polyvinylidene fluoride (PVDF). For example, Contestabile [35] developed a laboratory process for recycling LIBs in which the anode and cathode that were obtained after removing the shell were heated with an NMP solution below 100 °C. This method allowed for LiCoO₂ and graphite to be separated from the collector effectively. Zhou [36] chose dimethylformamide (DMF) to dissolve PVDF. They showed that the solubility of PVDF in DMF at 60 °C is 176 g/L. However, the solvents used in the separation process are very expensive and have a certain degree of toxicity, thus posing a threat to the environment and human health. In addition, the dissolved residues still adhere to the surface of electrode material particles, which are difficult to effectively remove, thereby affecting the flotation efficiency of the electrode materials. The pyrolysis method has the advantage of low secondary pollution. Sun and Qiu [37,38,39,40] proposed a novel method with which to separate cathode materials by means of vacuum pyrolysis. Through the process of pyrolysis, the electrolyte and binder were evaporated or decomposed, which reduced the adhesion of the cathode material and collector. When the temperature was between 500 and 600 °C, the separation efficiency increased with the increase in pyrolysis temperature. However, vacuum pyrolysis and pyrolysis under an inert atmosphere are high-cost shortcomings and pose operational difficulties, posing further difficulties with respect to their application in industrial production. Therefore, a low-cost but efficient method of removing the binder PVDF and recovering the graphites and cathode materials from spent LIBs is urgently needed. Roasting–froth flotation could be an already-available method with which to achieve this purpose. The process of roasting under an air atmosphere is adopted for the removal of the binder PVDF with industrialization feasibility. This method is simple and easy to perform. By controlling the roasting temperature, roasting can not only remove the binder PVDF from the spent LIBs but also reduce the loss of graphites. It is expected that the method of roasting–froth flotation will realize the efficient separation and recovery of graphites and cathode materials from spent LIBs.

In this study, the roasting–flotation method is used to remove the organic binder PVDF and separate the graphites and cathode materials from spent LIB materials. The roasting characteristics of the spent LIB materials are carefully analyzed. Thermogravimetric analysis is used to analyze the thermostability of the raw, spent LIB materials roasted in the air. The surface morphology and surface chemical states of the electrode materials under different roasting conditions are determined by SEM-EDS and XPS to obtain the optimum condition of roasting in air, and the mechanism of flotation separation enhanced by roasting is investigated. Based on the analysis results, froth flotation experiments of the spent LIB materials roasting at the optimal roasting condition are conducted to recover graphites and cathode materials and evaluate the separation feasibility of graphite from cathode materials via roasting–froth flotation.

2. Experimental

2.1. Materials

The spent LIBs used in this study are supplied by the local electronic waste center and they are discharged, crushed, and screened to obtain the spent LIB materials. Powder below 0.25 mm is chosen as the raw spent LIB material in this research. The phase composition of the raw spent LIB materials is determined via X-ray diffraction (XRD, Bruker D8 advance, Bruker AXS GmbH, Karlsruhe, Germany), and the setting conditions for the XRD are Cu Ka radiation (λ = 0.1541 nm), 40 keV accelerating voltage, 40 mA current, 3–90° scanning range, and 0.1 s/step (0.02°/step) scan speed. The XRD pattern of the spent LIB materials is shown in Figure 1. It is found from Figure 1 that the raw spent LIB materials are mostly composed of lithium nickel cobalt manganese oxide (NCM) and graphite. The metal element compositions of the raw spent LIB materials are analyzed by an inductively coupled plasma optical emission spectrometer (ICP-5000R, Focused Photonics Inc., Beijng, China) and the results are given in

Table 1. Metal element compositions of the raw spent LIB materials.

Elements	Li	Ni	Co	Mn	Others (Graphite + Organics)
Wt/%	3.28	8.82	7.14	13.05	67.71

. From Table 1, it can be seen the metal elements in the raw spent LIB materials are Ni, Co, Mn, and Li. The contents of Ni, Co, Mn, and Li are 8.82%, 7.14%, 13.05%, and 3.28%, respectively. From Figure 2, it can be seen that the particle size of the raw spent LIB powders ranges from 6 to ~60 μm, indicating that the raw spent LIB materials have a fine particle size.

2.2. Flotation Experiments

An XFG-type laboratory flotation machine (Jilin Exploration Machinery Plant, Jilin, China) with a volume of 200 mL is used to float the spent LIB materials and the impeller rotation speed is fixed at 1900 r/min. The XFG-type laboratory flotation machine has several flotation cells of different volumes from 20 mL to 200 mL for different amounts of ores from 5 g to 50 g. Using this machine to float a 10 g sample is quite suitable. In the initial phase, about 100 mL of deionized water is used to moisten 10 g of spent LIB material samples and the pulp is transferred to the flotation cell. Subsequently, deionized water is added to the marker line of 200 mL of the flotation cell. Then, the collector (kerosene) is subsequently added to the mineral suspensions and agitated for 2 min. Finally, methyl isobutyl carbinol (MIBC) is added as a frother and the flotation flowsheet is shown in Figure 3. The recovery of graphite is calculated after 3 min of flotation, after which concentrates (graphites) and tailings (cathode materials) are collected, dried, and weighted to calculate the flotation recovery, which can be calculated using the following equation:

$$\mathrm{R} = \frac{m_a}{m_b} \times 100\%$$

For graphite, where m_a stands for the weight of graphite in concentrates (g), m_b stands for the total weight of graphite in the flotation feed (g), and R is the recovery (%). For cathode materials, where m_a stands for the weight of cathode materials in tailings (g), m_b stands for the total weight of cathode materials in the flotation feed (g), and R is the recovery (%).

2.3. Analytical Methods

Thermogravimetric analysis of the raw spent LIB materials is conducted using a thermogravimetric analyzer (Discovery, TGA) over a temperature range of 25–850 °C at a heating rate of 10 °C/min under an air atmosphere. Meanwhile, the airflow is fixed at 40 mL/min. The spent LIB materials before and after roasting are analyzed with a scanning electron microscope (SEM, ZEISS Gemini 300, Carl Zeiss Co., Ltd, Oberkochen, Germany) and X-ray photoelectron spectrometer (XPS, Thermo Scientific K-Alpha, Thermo Fisher Scientific, Waltham, MA, USA) to show the surface properties. Scattered electron (SE) images of the samples are obtained via SEM and coupled with an energy-dispersive spectrometer (EDS) detector to allow elemental mapping. The analysis is performed in a low-vacuum pressurized chamber (10 Pa). A 15 kV electron beam and a working distance of around 8.5 mm are used. The XPS tests are carried out at room temperature in an ultra-high-vacuum (UHV) system. The base pressure of the analysis chamber during the measurements is lower than 3 × 10⁻⁷ mbar. Al Ka radiation (hv = 1486.6 eV) is from a monochromatized X-ray source. For all analyses, the take-off angle of the photoelectrons is 90° and the spot size is 400 μm. High-resolution spectra are recorded with a pass energy of 50 eV, and the energy step size is 0.1 eV. For flotation concentrate and tailings, 0.1 g is accurately sampled each time and dissolved in 6 mL aqua regia, and then each metal content is tested via ICP-OES to calculate metal recovery and grade.

3. Results and Discussion

3.1. Surface Morphology Analysis of Spent LIB Materials before Roasting

First, a clear investigation of the surface properties of raw spent LIB materials is of great significance to understanding the reason why they need to be roasted before froth flotation. SEM-EDS is used to analyze the surface morphology characteristics of the raw spent LIB materials and the results are shown in Figure 4. Figure 4a,b show the SEM images of raw spent LIB materials and EDS images are shown in Figure 4c–h. Obviously, the raw spent LIB materials are mainly composed of F, C, O, Ni, Co, and Mn elements. It is known from the previous literature [31] that the F element is mainly from PVDF binders and lithium hexafluorophosphate (LiPF₆) electrolytes, and the C element is mainly from graphite. In addition, the O, Ni, Co, and Mn elements are mainly from cathode materials. Based on the distribution of the four elements C, Ni, Co, and Mn on the surface of the raw spent LIB materials, combined with the previous XRD analysis, it is known that the smaller particles are cathode materials (NCM), and the larger particles and the lamellar materials are anode materials (graphites), as shown in Figure 4a,b. Some floccules (which are mainly made up of F, O, and C elements) can be clearly seen on the surface of graphites and cathode materials. It can be inferred that the floccules are mainly from the PVDF binder and other organics. The PVDF binder in the raw spent LIB materials is evenly distributed on the surface of the electrode materials and closely bonded with the graphite and cathode materials. It is difficult to separate graphites and cathode materials. Therefore, the effective removal of the PVDF binder is important for the following flotation separation of graphites and cathode materials from spent LIB materials.

3.2. Thermogravimetric Analysis of Raw Spent LIBs Materials

To accurately obtain the thermogravimetric characteristics of the raw spent LIB materials, the thermogravimetric analysis of raw spent LIB materials was carried out. Figure 5 shows the thermogravimetric analysis results of the raw spent LIB materials. From Figure 5, the spent LIB materials present four weightlessness stages, all in the range of room temperature to 850 °C. The first weight-loss stage occurs in a temperature range of room temperature to 150 °C. At this stage, the weight loss rate is 4.08%, and there is no obvious endothermic or exothermic peak in the derivative thermogravimetric (DTG) curve, which is attributed to the water being volatilized [38]. The second weight-loss stage appears in the temperature range of 150–450 °C. This is attributed to the thermal decomposition of electrolytes remaining in the spent LIB materials [39], and the weight loss rate is 5.18%. The third weight-loss stage is in the temperature range of 450–550 °C. This stage mainly comprises the combustion reaction of PVDF on the surface of the electrode materials [37], and the weight loss rate is 3.48%. The last weight-loss stage is in the temperature range of 550–850 °C. At this stage, the graphite is mainly oxidized and burned, and the weight loss rate reaches 28.41% [40]. Overall, the decomposition temperature of the binder PVDF is between 450 and 550 °C and the optimal roasting temperature of the spent LIB materials needs to be explored in greater detail.

3.3. SEM-EDS Analysis of Spent LIB Materials at Different Roasting Temperatures

To intuitively observe the surface morphology changes and element characteristics of spent LIB materials under different roasting conditions, the SEM-EDS analysis is used to investigate the spent LIB materials after roasting at different temperatures for an hour, and the results are shown in Figure 6. As can be seen from Figure 6a–d, after the spent LIB materials are roasted at 400 °C for 1 h, the range of the F element is significantly reduced and the floccule (the PVDF binder) still exists on its surface, indicating that the PVDF binder on the surface of the spent LIB materials is partly removed. As the roasting temperature increases, when the roasting temperature reaches 500 °C, the floccule on the surface of the electrode materials disappears, indicating that the binder PVDF on the surface of the spent LIB materials is completely removed. In addition, the EDS results suggest that the F element is uniformly distributed on the surface of the cathode materials. When the roasting temperature reaches 600 °C, as can be seen from Figure 6i–l, there is also no PVDF binder in the spent LIB materials. The surfaces of the graphites in Figure 6j are rougher than in Figure 6f. This is because most of the graphites had been oxidized when the roasting temperature reached 600 °C. The EDS analysis shows that F elements still appear on the surface of the cathode material, indicating that the decomposition product of the PVDF binder can easily react with the valuable metals (Ni, Co, Mn) to produce corresponding metal fluoride. Combining Figure 4 with Figure 6, it is deducted that roasting could remove the PVDF binder from the surfaces of spent LIB materials. As the roasting temperature increases, the PVDF binder and electrolytes are gradually removed, but too high a temperature could lead to the oxidation or even the burning of graphites. The optimal roasting condition to completely remove the organic binder PVDF while avoiding the loss of graphites is roasting at 500 °C for 1 h. To sum up, under the optimal roasting condition, the binder PVDF and electrolytes in the spent LIB materials are completely removed so that the cathode materials and graphites are effectively separated, resulting in a significant increase in the floatability difference between the cathode materials and graphites, which improves the feasibility of flotation.

3.4. XPS Analysis

XPS is a useful tool to detect the surface chemical states of active materials [41]. Herein, XPS is used to analyze the surface properties of spent LIB materials before and after roasting. Figure 7 shows the C spectra of spent LIB materials before and after roasting at the optimum roasting condition, and the peak binding energy and atomic content of C are shown in

Table 2. XPS spectrum composition of C1s before and after roasting.

Components	Raw Spent LIB Materials			Components	Roasted Materials
	Peak (BE)	FWHM (eV)	Atomic (%)		Peak (BE)	FWHM (eV)	Atomic (%)
Graphite	284.4	0.93	43.05	Graphite	284.4	0.68	50.55
C-C/C-H	284.8	1.73	27.53	C-C/C-H	284.8	1.64	25.00
C-COOR	285.7	1.21	13.68	C-H	286.6	2.40	12.60
-(CH2CF2)-n	286.6	0.93	5.45	C=O	288.8	2.40	5.92
O-C-O	287.8	1.10	3.62	C-F	291.6	2.40	5.92
O-C=O	288.5	1.33	3.09
-(CF2CH2)-n	290.5	1.31	3.44

.

From Figure 7a, before roasting, the C 1s are composed of seven peaks with binding energies of 284.4, 284.8, 285.7, 286.6, 287.8, 288.5, and 290.5 eV. The binding energy of 284.4 eV is mainly graphite material with a content of 43.05%. The binding energy of 285.7, 287.8, and 288.5 eV mainly contribute to C-COOR, O-C-O, and O-C=O, respectively. In addition, the functional group -(CH₂CF₂)-n at 286.6 eV and -(CF₂CH₂)-n at 290.5 eV are mainly derived from the organic PVDF binder on the electrode surface [42]. Figure 7b presents the C spectra of spent LIB materials after roasting. From Figure 7b, the C 1s are composed of five peaks with binding energies of 284.4, 284.8, 286.7, 288.6, and 291.6 eV. The C1s peak at 284.4 eV was mainly graphite, and its content increased from 43.05% before roasting to 50.55%, indicating that organics in the electrode material were effectively removed and more of the graphite surface was exposed. The binding energy at 288.8 eV is mainly contributed by C=O, which is mainly caused by the thermal decomposition and oxidation of C-COOR, O-C-O, and O-C=O in the organics. The binding energy of 286.6 and 291.6 eV is mainly contributed to C-H and C-F, which is mainly caused by the thermal decomposition of PVDF. Overall, the organic PVDF binder in spent LIB materials is effectively removed at the optimum roasting condition.

Figure 8 shows the F spectra of spent LIB materials before and after roasting at the optimum roasting condition, and the peak binding energy and atomic content of F are shown in

Table 3. XPS spectrum composition of F1s before and after roasting.

Components	Raw Spent LIB Materials			Components	Roasted Materials
	Peak (BE)	FWHM (eV)	Atomic (%)		Peak (BE)	FWHM (eV)	Atomic (%)
Li-F	685.67	1.41	23.43	Li-F	685.10	1.57	29.17
Al-F	686.97	1.69	29.56	Mn-F/Ni-F	685.90	1.45	60.87
-(CH2CF2)-n	687.85	1.63	33.27	Al-F/Co-F	686.89	1.61	6.91
P-F	688.52	2.40	13.84	C-F	687.85	1.44	3.05

. From Figure 8a, before roasting, the F 1 s are composed of four peaks with binding energies of 685.67, 686.97, 687.85, and 688.52 eV, attributed to Li-F, Al-F, -(CH₂CF₂)-n, and P-F. The P-F peak at 688.52 eV is derived from the electrolyte (LiPF₆) [43]. The binding energy of -(CH₂CF₂)-n at 687.85 eV is derived from PVDF, and the atomic ratio is 33.27%. The Al-F with binding energy at 686.97 eV is mainly generated in the charging and discharging process of the battery, and the content is 29.56%. Li-F with binding energy at 685.67 eV is derived from LiF compounds generated in the charging and discharging process of the battery and LiF compounds generated in the decomposition of LiPF₆, and account for 23.43%. After roasting, from Figure 8b, the P-F peak disappeared completely, suggesting that the LiPF₆ electrolyte underwent complete decomposition. The C-F peak at 687.85 eV is attributed to the oxidative decomposition of PVDF, and the content is only 3.05%. The contents of the Al-F/Co-F peak and Mn-F/Ni-F peak at 686.69 eV and 685.90 eV are 6.91% and 60.87%, respectively, and are mainly derived from the transfer of the F element in electrolytes and organics and its combination with valuable metals such as Ni, Co, and Mn to form metal fluoride. The content of the Li-F peak at 685.10 eV is 29.17%, which is partly due to the transfer of the F element in electrolyte and organic binder and its combination with valuable metal Li to form metal fluoride. According to the analysis of the chemical state of the F element combined with the existing literature, a possible reaction route was inferred (Reactions (1)–(3)) [39,43]. All in all, the XPS results suggest that the organic binders and electrolytes have been completely decomposed after roasting. Parts of the valuable metals react with the decomposition products of organics to produce corresponding metal fluoride, and the results are consistent with SEM analysis.

LiPF₆ → LiF + PF₅

(1)

PF₅ + H₂O → POF₃ +2HF

(2)

POF₃ + H₂O →POF₂(OH) +HF

(3)

3.5. Flotation Experiments

The cathode materials of roasted spent LIB materials are mainly composed of NCM, which has strong hydrophilicity. In addition, the anode materials are composed of graphites, which have strong hydrophobicity [44]. Therefore, froth flotation technology can be used to effectively separate graphites and cathode materials. According to the results of the existing literature, using kerosene as a collector and MIBC as a frother presents an excellent collecting effect on graphite [14]. Therefore, in this manuscript, the effects of kerosene and MIBC concentrations on the separation of graphites and cathode materials in the spent LIB materials after roasting at 500 °C for 1 h are discussed, respectively, and the results are shown in Figure 9 and Figure 10. Among them, Figure 9a and Figure 10a present the grades and recoveries of graphites in concentrates, while Figure 9b and Figure 10b display the grades and recoveries of cathode materials in tailings.

Figure 9 shows the effect of different kerosene dosages on flotation results. From Figure 9, it can be noted that the kerosene displays an excellent collecting ability for graphites when the MIBC dosage is fixed at 100 g/t, indicating that the kerosene can be used to separate graphite from spent LIB materials. The grade of graphites and cathode materials gradually increased and then became stable with the increase in kerosene dosage. Meanwhile, the recoveries of graphites and cathode materials gradually decreased. When the kerosene dosage reaches 200 g/t, the cathode materials’ grade is 91.32% with a recovery of 92.93%, and the graphite grade is 85.02% with a recovery of 81.96%. In addition, with the continued increase in kerosene dosage, the recoveries of graphites and cathode materials remain almost constant. Thus, the optimal kerosene dosage is fixed at 200 g/t. Based on this, the effect of MIBC dosage on the flotability of graphites is further investigated, and the results are shown in Figure 10. From Figure 10, with an increase in MIBC dosage, the recovery of graphite increases slowly from 74.9% at the dosage of 25 g/t to 82.7% at the dosage of 150 g/t. Meanwhile, the grade of cathode materials increases gradually from 88.3% to 91.6%. Furthermore, an increase in MIBC dosage results in a slow decrease in graphite recovery and cathode material grade. When the dosage of MIBC increases from 150 g/t to 250 g/t, the recovery of graphite decreases steadily from 82.7% to 81.1%, and the grade of the cathode materials decreases from 91.6% to 91%. It is apparent that excessive MIBC will reduce the separation efficiency of graphite and cathode materials. Importantly, the grade and recovery of both graphites and cathode materials are high during the whole flotation process, which indicates that it is easy to separate graphites from cathode materials when the spent LIB materials are roasted at 500 °C for 1 h. In conclusion, when the kerosene dosage reaches 200 g/t and the MIBC dosage reaches 150 g/t, the best separation of graphites and cathode materials is reached: the cathode material grade is 91.6% with a recovery of 92.6%, and the graphite grade is 84.6% with a recovery of 82.7%.

4. Conclusions

The recovery of graphites and cathode materials represents a strategic point in the recycling process of spent LIB materials. In this paper, a roasting–froth flotation process for recovering the graphite and cathode materials of spent LIBs is investigated. In the range of 400–600 °C, the PVDF binder and electrolytes on the surface of the spent LIB materials are gradually removed with an increase in roasting temperature, and the optimal roasting temperature and time are 500 °C and 1 h, respectively. The organic PVDF binder on the surface of the spent LIB materials is effectively removed and the graphites show no obvious loss under optimal conditions. After roasting, the cathode materials and graphites are obviously separated from the spent LIB materials, which greatly improves the floatability difference between cathode materials and graphites. A comprehensive investigation and evaluation of the recovery process of spent LIB materials are undertaken based on the analysis roasting process. The flotation results suggest that it is easy to separate graphites from cathode materials when the spent LIB materials are roasted at the optimal roasting condition. When the dosage of kerosene reaches 200 g/t and the dosage of MIBC is 150 g/t, the cathode materials grade is 91.6% with a recovery of 92.6% and the graphite grade is 84.6% with a recovery of 82.7%. Based on this present result, graphites and cathode materials can be easily recovered through roasting–froth flotation. Then, via further purification operations, high-quality graphites and cathode material products can be obtained without losing the graphites from the spent LIB materials. Thus, this may lay the foundation for closing the loop for spent LIB materials in the future.