Pretreatment Methods for Recovering Active Cathode Material from Spent Lithium-Ion Batteries

Abstract

The development of environmentally friendly pretreatment processes for spent lithium-ion batteries (LIBs) is crucial for optimizing direct recycling methods. This study explores alternative approaches for recovering active cathode materials from end-of-life LIBs, focusing on environmentally safer options compared to the usually employed toxic solvent N-methyl-pyrrolidone (NMP), using disassembled batteries as test subjects. Various pretreatment methods, including thermal treatment, selective aluminum foil dissolution with a NaOH solution, and the use of eco-friendly solvents such as triethyl phosphate (TEP), are examined on the cathode sheets. The results show that thermal pretreatment combined with TEP provides the most effective approach, achieving a recovery efficiency of 95% while maintaining the morphology and purity of the recovered materials, making them suitable for direct recycling. These methods are further tested on complete battery cells, simulating industrial-scale operations. The TEP treatment proves particularly promising, ensuring high recovery efficiency and preserving the structural integrity of the materials, with a mean particle diameter of approximately 8 µm. Additionally, when applied to cycled batteries, this pretreatment successfully recovers active materials without contamination. This study provides valuable insights into various pretreatment strategies, contributing to the development of a greener, more efficient direct recycling pretreatment process for spent LIBs.

1. Introduction

In the current global landscape, the increase in carbon dioxide (CO₂) emissions represents an urgent and critical challenge for environmental sustainability. The atmospheric concentration of CO₂ has risen from 280 ppm before the Industrial Revolution to approximately 420 ppm today, an increase of nearly 50% [1,2]. CO₂ is one of the main greenhouse gases, and its continuous emission and persistence in the atmosphere are among the primary causes of climate change affecting our planet. During the 2015 Paris Conference of the Parties (COP), a concrete target was identified for the first time: to limit the temperature increase to below 1.5 °C compared to pre-industrial levels [3]. Addressing this issue requires innovative strategies across multiple sectors, including energy, transportation, and resource management [4,5,6].

Lithium-ion batteries (LIBs) have emerged as a key technology for decarbonizing the energy and transport sectors. LIBs are critical to reducing reliance on fossil fuels by enabling the transition to electric vehicles (EVs) and renewable energy systems [7,8]. Today, nearly all commercially available batteries use lithium-ion technology, which offers several advantages, such as high energy density (150–250 Wh/kg), high charge/discharge efficiency (90–95%), long cycle life (1000 to 3000 charge/discharge cycles), and reduced weight [9,10]. However, the widespread adoption of LIBs also presents significant challenges, particularly in terms of raw material supply, environmental impact, and end-of-life management [11]. Additionally, commonly used materials in LIBs, such as lithium, cobalt, and graphite, are currently included on the EU’s list of critical raw materials (CRMs) due to their strategic importance and high supply risks caused by geopolitical limitations and extraction difficulties [12]. The global demand for LIBs is expected to grow exponentially, driven by the electrification of transport and the expansion of renewable energy infrastructure [13]. The EV market is projected to exceed USD 725 billion by 2026, with the peak annual battery production expected around 2040, reaching over 4 million tons, excluding batteries for other applications like smartphones and laptops [14]. The International Energy Agency (IEA) estimated that, in 2019, LIB waste from EVs amounted to approximately 500,000 tons, a figure that could rise to 8 million tons by 2040 [11]. In 2016, 78,000 tons of lithium carbonate (Li₂CO₃) were consumed, a 2.7-fold increase from 2010. Annual global demand is expected to reach 509,000 tons of Li₂CO₃ by 2025. In 2017, 35% of extracted lithium was used for LIBs, a figure projected to rise to 66% by 2025 [15]. Most end-of-life LIBs whose capacity has dropped below 80% of the original are currently landfilled rather than recycled [16]. This practice not only necessitates the continuous use of newly extracted materials but also causes significant environmental problems, as the electrolyte solution can leak from LIBs, contaminating groundwater and harming ecosystems [11,17].

Recycling precious metals, however, offers environmental benefits and can ensure a circular flow of critical battery elements, reducing dependency on other nations. In Europe, less than 5% of LIBs were recycled in 2022, with only 1% of lithium recovered [11]. Given these figures, scientific research is urgently needed to ensure the economic and technical feasibility of industrial-scale recycling methods for recovering valuable materials from cathodes and anodes, enabling a circular economy. Commonly, LIBs comprise a cathode, an anode, a separator, an electrolyte solution, current collectors (like aluminum and copper foils), a protective shell, and a container [18,19]. Conventional recycling methods include pyrometallurgy and hydrometallurgy, which primarily focus on reintegrating recovered high-value materials into the production of new LIBs [20]. Despite their widespread industrial application, these processes currently face significant challenges, such as the high energy intensity associated with pyrometallurgy and the substantial wastewater generation associated with hydrometallurgy [11,20,21]. In contrast, direct recycling has generated significant attention due to the possibility of recovering the cathode materials without breaking down their crystal structure [22], potentially offering a more affordable process that requires less energy [23]. A critical step in this process is the separation of the active cathode material from the Al foil by breaking the bond between the polyvinylidene fluoride (PVDF) binder and the active material [14]. The most commonly used organic solvents are N,N-dimethylformamide (DMF), N-methyl-pyrrolidone (NMP), and N,N-Dimethylacetamide (DMAC). Among them, the NMP is largely utilized due to its ability to dissolve the PVDF, thereby stripping the cathode material [24,25]. Still, those solvents are classified as toxic and carcinogenic [14,26]. To address the limitations, recent studies have explored the use of greener alternatives, such as triethyl phosphate (TEP), Cyrene, $\gamma$-valerolactone (GVL), and dimethyl isosorbide (DMI), which have demonstrated high efficiency in recovering spent cathode materials without introducing impurities [14,27]. An alternative approach involves using a NaOH solution to selectively dissolve the aluminum foil, achieving high separation efficiency [28]. Another strategy involves the thermal decomposition of the binder at elevated temperatures, allowing the separation of the active cathode material from the metal collectors [29]. Moreover, preserving the morphology of the recovered material is essential to ensure an optimal regeneration process from the direct recycling perspective [30].

Typically, previous studies have focused exclusively on a single pretreatment approach, be it thermal or solvent-based. In contrast, our study simultaneously investigates both pretreatment methods using the same initial battery material. This comprehensive approach allows us to directly compare the methodologies and determine which one is the most promising route from a direct recycling point of view. In particular, three approaches are examined and compared to the NMP pretreatment: thermal treatment, the selective dissolution of aluminum foil using a NaOH solution, and the application of eco-friendly solvents such as triethyl phosphate (TEP). Initially, this study focuses on the cathode sheet to identify the most effective recovery process. The evaluation considers key factors such as yield, recovery efficiency, impurity levels, stoichiometry, and the morphological integrity of the recovered cathode material. Subsequently, the analysis is extended to the entire battery cell to assess the best-performing methods in a more realistic industrial context. Finally, the most effective methodology is applied to a cycled battery to evaluate its practical performance. Our work seeks to build a broader knowledge of pretreatment efficacy, ultimately guiding the development of more effective direct recycling processes.

2. Materials and Methods

2.1. Materials

In this study, Samsung SDI INR 18650-20R batteries (Samsung Electronics Co., Ltd., Suwon, Republic of Korea) were used. The cathode comprised an aluminum foil current collector, polyvinylidene fluoride (PVDF) binder, carbon black (CB), and the active material LiNi_0.5Mn_0.3Co_0.2O₂ (NMC532), while the anode consisted of a copper foil current collector, graphite, and binder. Dimethyl carbonate (DMC, >99%, Sigma-Aldrich, St. Louis, MO, USA) was employed to remove the electrolyte. For cathode material recovery, sodium hydroxide (NaOH, >98%, Sigma-Aldrich), N-methylpyrrolidone (NMP, >99.5%, Sigma-Aldrich), triethyl phosphate (TEP, >98%, ThermoScientific, Waltham, MA, USA), and bi-distilled water were used.

2.2. Active Material Recovery Process

The batteries were discharged due to safety reasons and disassembled (see Figures S1–S3 of the Supplementary Materials for further details), and, then, the electrodes were washed with DMC to remove the electrolyte. Thermal treatment, aluminum foil selective dissolution using a NaOH solution, and organic solvent (NMP and TEP) treatment were investigated to recover the active cathode material. The schematic representation of the applied pretreatments to the discharged batteries are reported in Figure 1 and Figure 2. Figure 1 depicts the experimental steps of the thermal treatment on the cathode sheet (Figure 1A) and on the entire battery (Figure 1B), while Figure 2 reports the solvent-based treatment phases on the cathode (Figure 2A) and on the entire cell (Figure 2B).

Initially, for all the tests, the cathode sheet was separated from the anode and the separator and subsequently cut into small pieces (1 × 1 cm²) for the recovery process.

For the thermal pretreatment, the electrode pieces were first treated at a high temperature in a muffle furnace for binder removal and then immersed in a bi-distilled water solution with magnetic stirring to facilitate the separation of the active material from the aluminum foil. The powder was then filtered, weighed, dried, and recovered. In this case, no additional treatments are required since the powder is already purified from the binder and carbon black due to the degradation occurring at the high temperatures applied. A schematic representation of the process is depicted in Figure 1A.

For solvent-based pretreatments, the operational procedures involved heating the solvents to the desired temperatures and then adding several pieces of cathode materials under magnetic stirring or ultrasonication to facilitate the separation of the active cathode material from the aluminum foil. The process steps for solvent-based pretreatments on the cathode sheet (using NaOH solution, NMP, or TEP) are reported in Figure 2A.

The aluminum foil dissolution was performed in a NaOH aqueous solution (10 wt% of NaOH), which was then filtered under vacuum to obtain the powder and subsequently dried and weighed. Then, the powder underwent a thermal treatment at high temperatures to remove the carbon black and binder residues. In the pretreatment involving NMP and TEP, the organic solvent facilitated binder dissolution, enabling electrode delamination. The aluminum foil was manually separated from the solution, and the powder was collected by vacuum filtration. Then, a thermal treatment was carried out to remove the carbon black and the PVDF residues.

Different operating conditions were tested to identify the optimal process parameters for each methodology. A detailed summary of all the experimental trials is provided in

Table 1. Operating conditions for the pretreatment of the cathode sheet.

Methodology	ID	Operational Conditions				Muffle Treatment
		T [°C]	Stirring/Ultrasonication	S/L Ratio [g/mL]	Time [h]	T [°C]
Thermal pretreatment	TH-1	25	400 rpm	1/30	2	570
	TH-2	25	400 rpm	1/30	4	570
	TH-3	25	300 rpm	1/30	2	720
NaOH leaching	NaOH-L1	70	500 rpm	1/10	2	570
	NaOH-L2	90	500 rpm	1/20	4	570
NMP pretreatment	NMP-1	100	300 rpm	1/10	1	800
	NMP-2	50	ultrasonication	1/10	3	800
TEP pretreatment	TEP-1	120	400 rpm	1/20	2	570

. Stirring rates in the 300–500 rpm range were chosen according to values typically reported in the literature [31,32,33]. A stirring rate below 300 rpm was not investigated since such a low mixing velocity could lead to poor recovery efficiency of the cathode material; whereas a stirring rate higher than 500 rpm was not considered as excessive mixing could cause particle damage, obtaining a degraded morphology. The temperature of the muffle treatment was selected to allow the thermal oxidation of cell components (e.g., carbon black and graphite). Furthermore, the solid–liquid (S/L) ratios employed in the pretreatment trials aligned with typical values reported in the literature [34,35], enabling a compromise between high recovery efficiencies and reduced solvent consumption.

Afterward, the most promising procedures were applied to the entire cell (comprising the cathode, anode, and separator) to simulate a real industrial application. In the thermal pretreatment process, the separator’s presence did not significantly impact the recovery process, while the presence of anodic graphite posed a challenge due to its high oxidation temperature. To prevent aluminum melting (660 °C) and minimize copper oxidation, the operating temperature was limited to 350 °C. Accordingly, a two-step treatment in the muffle furnace was required to obtain the active cathode material free of impurities. In this case, after the first thermal treatment, the electrode pieces were immersed in a bi-distilled water solution and separated from the Al and Cu foils using magnetic stirring or ultrasonication. After being recovered by vacuum filtration, the powder underwent a second thermal treatment at a higher temperature (720 °C) to decompose the anodic graphite and the carbon black present in the material. A thermal pretreatment in an inert atmosphere was also investigated using pure nitrogen to avoid copper oxidation at high temperatures. These pretreatments are depicted in Figure 1B.

The recovery of the active material from the entire battery was also performed using TEP, selected for its non-toxic nature and the excellent performance observed in a similar procedure applied to the cathode sheet (see Figure 2B). The experimental tests with TEP were performed in two replicates (TEPC-1 and TEPC-1R) to confirm the reproducibility of the favorable results and on a fully cycled battery (TEPCC) to assess the pretreatment performance under more realistic conditions. A detailed summary of the experimental conditions on the entire battery is reported in

Table 2. Operating conditions for the pretreatment of the entire cell.

Methodology	ID	Operational Conditions				Muffle Treatment
		T [°C]	Stirring/Ultrasonication	S/L Ratio [g/mL]	Time [h]	T [°C]
						1°	2°
Thermal Pretreatment (air)	THCa-1	25	400 rpm	1/30	4	350	720
	THCa-2	25	ultrasonication	1/30	3	350	720
	THCa-3	25	ultrasonication	1/30	1	350	720
Thermal Pretreatment (N2)	THCi-1	25	500 rpm	1/30	4	350	720
TEP Pretreatment	TEPC-1	120	400 rpm	1/20	2	/	720
	TEPC-1R	120	400 rpm	1/20	2	/	720
	TEPCC	120	400 rpm	1/20	2	/	720

.

2.3. Material Characterization

The recovered active material was analyzed with various techniques to assess its purity, chemical composition, crystal structure, and surface morphology. The elemental composition of the cathode material was determined by inductively coupled plasma optical emission spectrometry (ICP-OES) using a Thermo iCAP 7400DUO spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) after sample digestion. Thermogravimetric analysis (TGA) was conducted using a Q500 analyzer (TA Instruments, New Castle, DE, USA) to determine the mass fraction of organic components (binder, carbon black, and graphite) contained in the cathodic and anodic sheets. Constant heating rate (10 °C/min) runs in air atmosphere were carried out. The amount of each component was determined by analyzing the characteristic oxidation peaks and the corresponding weight loss. Powder X-ray diffraction (XRD) was used to investigate the crystal structure of the recovered material and to compare the characteristic peaks with the one of the pristine NMC532 available in the literature. The diffractometer employed was a Bruker D2 Phaser (Bruker Corporation, Billerica, MA, USA). The analysis was conducted on a 2θ range between 10° and 80°. Furthermore, the powder morphology was analyzed through scanning electron microscopy (SEM) using a FEI Quanta FEG 450 (FEI Inc., Hillsboro, OR, USA). The software ImageJ2 was used to estimate the mean particle diameter from the SEM analysis.

2.4. Performance Indices of the Pretreatment

Different parameters were defined to compare the performance of each method used for the active material recovery. Equation (1) reports the yield ($\mu$) of the recovered cathode material:

$$\mu \left[\%\right]={\displaystyle \frac{m_r}{m_b }}·100$$

(1)

where $m_r$ represents the mass of the cathode material recovered from the pretreatment, and $m_b$ is the initial battery mass without the external cover. A recovery efficiency (${\eta }_{NMC}$) was then defined according to Equation (2) as the ratio between the recovered material and the theoretical maximum quantity that can be recovered.

$${\eta }_{NMC} \left[\%\right]={\displaystyle \frac{m_r}{m_t }}·100$$

(2)

where $m_t$, which is the theoretical mass of cathode material that can be recovered, can be expressed using Equation (3):

$$m_t=m_c·\left(1-x_{al}-x_{PVDF}-x_{CB}\right)$$

(3)

where $m_c$ is the cathode mass, $x_{al}$ is the fraction of aluminum, $x_{PVDF}$ is the fraction of binder, and $x_{CB}$ is the fraction of carbon black in the cathode. Furthermore, the recovery efficiency of the single elements was defined as the ratio between the recovered mass of a certain element and the theoretical quantity that can be recovered of the same element according to Equation (4):

$${\eta }_i \left[\%\right]={\displaystyle \frac{m_r·w_i}{m_c·x_i}}·100$$

(4)

where $w_i$ is the mass fraction of the recovered element $i$ (Li, Ni, Mn, Co), and $x_i$ is the mass fraction of the same element in the cathode obtained from the preliminary tests conducted on the batteries. $w_i$ and $x_i$ were determined by ICP-OES analysis. Lastly, to assess the stoichiometry of the recovered material, the molar ratio is calculated as follows:

$$f_i={\displaystyle \frac{n_i}{n_{Ni}+n_{Mn}+n_{Co}}}$$

(5)

where $n_i$ are the moles of the $i$ component (Ni, Mn, Co, Li), $n_{Ni}$ are the moles of nickel, $n_{Mn}$ are the moles of manganese, and $n_{Co}$ are the moles of cobalt.

3. Results and Discussion

Preliminary tests were conducted on the battery cathode and anode to assess the binder fraction, the carbon black fraction, and the current collector fraction and to confirm the battery chemistry. From the TGA tests on the cathode, the weight loss up to 450 °C was associated with the oxidation of the binder, while the weight loss in the 450–550 °C range was associated with carbon black (see Figure S4 of the Supplementary Materials). Overall, the TGA results indicated a total fraction of the organic binder and carbon black of approximately 4.5 wt%. The aluminum mass fraction in the cathode sheet was determined using ICP-OES analysis. This technique enabled the quantification of the mass fractions of various elements in the cathode, as well as the stoichiometric composition of the cathode active material (see Tables S1 and S2 in the Supplementary Materials, respectively). A summary of the cathode composition is reported in Figure 3. The TGA test on the anode highlighted the oxidation temperature of the anodic graphite (Figure S5 of the Supplementary Materials). SEM images of the cathode and anode sheets are shown in Figures S6 and S7 of the Supplementary Materials.

3.1. Pretreatment on the Cathode Sheet

Figure 4 depicts the recovery efficiency for the various methods applied only to the cathode sheet listed in Table 1. By analyzing this parameter, it is clearly visible that the three pretreatment processes that have the higher recovery efficiency are the NaOH solution leaching with a recovery efficiency of 98% (NaOH-L1), the thermal pretreatment with a recovery efficiency of 94%, and the pretreatment using TEP with a value of 97%. The NMP solvent shows low recovery efficiency, probably due to the relatively low temperature used in the dissolution step.

Overall, TEP pretreatment demonstrated a high recovery rate of pure cathode active materials, making it a promising option for direct recycling. The recovery efficiency (97%) achieved in this study surpasses the values reported in previous research. Rahman et al. [14] documented recovery efficiencies ranging from approximately 70% to 85% using TEP, with all tested green solvents yielding efficiencies below 90%. Similarly, Ahuis et al. [27] reported a maximum recovery yield of 69% with TEP. In addition, it is important to remark that process efficiency is strongly influenced by various operating parameters, including heating temperature, treatment duration and method (mechanical/magnetic stirring or ultrasonication), and the solid-to-liquid (S/L) ratio. In our experiments, we adopted a temperature of 120 °C, magnetic stirring for 2 h, and an S/L ratio of 1:20. This ratio falls within the range reported in previous studies, which typically use S/L ratios between 1:15 and 1:30 [14].

ICP-OES analysis was used to assess the recovery efficiency of transition metals and lithium (

Table 3. Yield and efficiency of elements recovery for the various pretreatments applied to the cathode sheet.

ID	$\mathit{\mu }$	${\mathit{\eta }}_{\mathit{N}\mathit{i}}$	${\mathit{\eta }}_{\mathit{M}\mathit{n}}$	${\mathit{\eta }}_{\mathit{C}\mathit{o}}$	${\mathit{\eta }}_{\mathit{L}\mathit{i}}$
TH-1	30.1	95.5	96.5	95.4	87.3
TH-2	27.7	91.1	89.8	89.4	79.7
TH-3	27.3	89.7	88.4	88.6	79.1
NaOH-L1	31.4	98.2	99.6	98.2	95.4
NaOH-L2	29.4	93.3	97.7	93.5	90.2
NMP-1	12.2	39.3	38.6	39.2	35.6
NMP-2	20.6	65.7	64.2	65.3	61.7
TEP-1	31.1	97	95.8	96.4	100

), to confirm that the stoichiometry remains unchanged (

Table 4. Results for the pretreatment on the cathode sheet from ICP-OES analysis.

ID	Weight Percentages in the Cathode Material [%]								${\mathit{f}}_{\mathit{i}}$ [mol/mol]
	Ni	Mn	Co	Li	Al	Fe	Zn	Cu	Ni	Mn	Co	Li
TH-1	28.59	16.84	11.94	6.08	0.21	0.03	0.20	0.30	0.489	0.308	0.203	0.879
TH-2	29.60	17.00	12.15	6.02	0.54	0.19	0.22	0.70	0.494	0.303	0.202	0.850
TH-3	29.48	16.94	12.18	6.05	0.40	0.19	0.22	0.70	0.494	0.303	0.203	0.857
NaOH-L1	28.16	16.66	11.78	6.37	1.14	0.05	0.17	0.22	0.488	0.308	0.203	0.933
NaOH-L2	28.59	17.45	11.98	6.43	0.13	0.15	0.36	0.15	0.483	0.315	0.202	0.919
NMP-1	29.11	16.65	12.15	6.13	0.09	0.08	0.96	0.00	0.493	0.301	0.205	0.879
NMP-2	28.72	16.35	11.94	6.27	0.28	0.38	1.34	0.00	0.494	0.301	0.205	0.912
TEP-1	28.06	16.16	11.66	6.94	0.03	0.10	0.23	0.50	0.493	0.303	0.204	1.03

), as well as to detect impurities (Table 4). The presence of impurities can significantly impact the electrochemical performance of the recovered materials. Their concentration varied depending on the pretreatment method employed. TEP yielded high-purity active materials (Table 4), reinforcing its suitability for direct recycling. These findings align with previous studies advocating the use of TEP for cathode pretreatment [14].

Table 4 presents the chemical compositions of the recovered NMC532 cathode material, obtained via ICP-OES analysis, under different treatment conditions. Compared to the untreated NMC532 cathode, the recovered cathode material consistently exhibited reduced lithium content, likely due to lithium leaching into the solution, with variations depending on the treatment conditions. However, the composition of transition metals (nickel, cobalt, and manganese) remained unchanged. As shown in Table 4, all the methods successfully recover the material without significantly altering the stoichiometry. The original 5:3:2 atomic ratio of Ni, Mn, and Co was preserved in all recovered materials, i.e., a consistent molar ratio closely matching that of the original cathode was obtained. Consequently, the recovered cathode active materials may be suitable for direct recycling.

Notably, only the TEP pretreatment yields a nearly stoichiometric quantity of lithium. In contrast, in all other cases, approximately 10% of lithium is lost during the process, probably due to lithium leaching in the solution. This is evident from the values reported in Table 3 for lithium recovery efficiency. Among the different methods investigated, the thermal pretreatment has the lower recovery efficiency of lithium (87%, as listed in Table 3), which means that almost 13% of Li is lost in the pretreatment compared with the initial quantity available. All the investigated pretreatment methods include an agitation step in water, except for those involving organic solvents, such as TEP pretreatment. In the presence of water, lithium leaching may occur, as suggested in the literature [27], leading to lower lithium recovery rates.

Preserving the morphology of cathode active materials during pretreatment is crucial for direct recycling procedures. Maintaining the structure of the aggregate of the primary particle is necessary to obtain reusable material. For this reason, the morphology of the recovered powders for each pretreatment was analyzed using SEM. The mean particle area and diameter were estimated from the SEM images, and the results are reported in

Table 5. Morphology results from the SEM analysis of the recovered powder for the various pretreatments on the cathode sheet.

ID	Mean Area [µm2]	Mean Diameter [µm]
Cathode ref.	56.2	8.5
TH-1	47.5	7.8
TH-2	18.9	4.9
TH-3	43.8	7.5
NaOH-L1	30.7	6.3
NaOH-L2	2.30	1.7
NMP-1	38.0	7.0
NMP-2	38.0	7.0
TEP-1	50.2	8.0

. From the SEM analysis (Figure S8 of the Supplementary Materials), it can be observed that the treatment NaOH-L1 has disintegrated the majority of the secondary aggregate. The mean particle diameter is about 6.3 µm, which is lower compared to the non-treated cathode. Furthermore, an undesirable precipitate that could be attributed to the crystallization of aluminum hydroxide is visible. The unwanted presence of Aluminum on the surface is also confirmed by the Energy-Dispersive Spectroscopy (EDS) analysis (Figure S9 of the Supplementary Materials). The treatment NaOH-L2, with a lower S/L ratio, higher temperature, and stirring velocity, exhibits a much lower mean particle diameter (1.7 µm). This could be the result of a more aggressive environment and a higher stirring velocity that could lead to severe damage to the particles. For the thermal pretreatment TH-1, the morphology of the recovered material appears intact and coherent with very few random damages. The estimated particle diameter is about 7.8 µm.

To further investigate the impact of temperature and stirring velocity, additional trials (TH-2 and TH-3) were conducted. The SEM images, shown in Figure S8 of the Supplementary Materials, indicate that increasing time and stirring velocity results in a more disaggregated structure with a smaller mean particle diameter of 4.9 µm. Conversely, temperature does not appear to influence particle morphology, as the mean particle size remained nearly unchanged between TH-1 and TH-3. In both recovery procedures using NMP, the morphology is maintained, showing only minor random damage, with a mean particle diameter around 7 µm. The most effective method for preserving the morphology of the recovered cathode material is pretreatment with TEP. This pretreatment ensures that the secondary aggregates remain unchanged, and the disaggregation phenomenon is negligible. The particle structure remains consistent, with a mean particle diameter of 8 µm. According to these results on the cathode sheet, for the direct recycling process, the most promising pretreatments are TH-1, TH-3, and the TEP pretreatment. These methods are then applied to an entire battery cell.

3.2. Pretreatment on the Entire Cell

Figure 5 and

Table 6. Yield and efficiency of element recovery for the various pretreatments applied to the entire cell.

ID	$\mathit{\mu }$	${\mathit{\eta }}_{\mathit{N}\mathit{i}}$	${\mathit{\eta }}_{\mathit{M}\mathit{n}}$	${\mathit{\eta }}_{\mathit{C}\mathit{o}}$	${\mathit{\eta }}_{\mathit{L}\mathit{i}}$
THCa-1	29.33	96.02	95.29	94.68	80.74
THCa-2	28.44	93.04	91.96	92.26	78.69
THCa-3	28.26	92.93	91.13	91.26	73.93
THCi-1	9.13	25.78	25.53	25.27	23.07
TEPC-1	27.87	90.35	89.46	89.09	86.62
TEPC-1R	27.31	86.61	85.09	84.76	79.45
TEPCC	28.90	93.49	91.18	92.98	87.00

show the results of the pretreatment methods applied to the entire cell (see Table 2). Except for the pretreatment THCi-1, all the methods exhibit a recovery efficiency higher than 85%. Comparing these results with those of the cathode alone, the efficiencies are lower due to the greater number of processing steps involved. To gain deeper insight into the different methods and assess process performance, further analyses were conducted to examine the purity and morphology of the recovered material.

The ICP-OES analyses presented in

Table 7. Results for the pretreatment on the entire cell from ICP-OES analysis of the recovered material.

ID	Weight Percentages in the Cathode Material [%]								${\mathit{f}}_{\mathit{i}}$ [mol/mol]
	Ni	Mn	Co	Li	Al	Fe	Zn	Cu	Ni	Mn	Co	Li
THCa-1	29.41	17.01	12.12	5.75	0.28	0.12	0.72	3.03	0.493	0.305	0.202	0.815
THCa-2	29.39	16.93	12.18	5.78	0.05	0.02	0.15	2.24	0.493	0.303	0.204	0.820
THCa-3	29.55	16.89	12.13	5.47	0.21	0.12	0.18	2.26	0.495	0.302	0.202	0.775
THCi-1	25.39	14.65	10.40	5.28	1.73	0.04	0.41	1.68	0.494	0.305	0.202	0.869
TEPC-1	29.13	16.81	12.01	6.49	0.05	0.02	0.14	0.95	0.493	0.304	0.203	0.930
TEPC-1R	28.50	16.32	11.66	6.08	0.06	0.02	0.23	1.36	0.495	0.303	0.202	0.893
TEPCC	29.07	16.53	12.09	6.29	0.07	0.01	0.22	1.69	0.495	0.300	0.205	0.905

clearly show that thermal pretreatment in air results in significant copper impurity content (2.25%) in the recovered material. In contrast, treatment in an inert atmosphere leads to a lower impurity level (1.68%), likely due to reduced copper oxidation. Still, THCi-1 has a recovery efficiency that is too low to be effectively considered for industrial application. Considering the stoichiometry of the recovered powder, almost 20% of lithium is lost in the thermal pretreatment. Pretreatment using TEP seems to be the most promising method. The copper impurity is around 1%. The lithium lost in the process is around 10%, which is half the quantity lost compared to the other methods.

The morphology of the recovered powders for each pretreatment was analyzed using SEM, and the obtained images are shown in Figure 6. The mean particle area and diameter were also estimated, and the results are reported in

Table 8. Morphology results from the SEM analysis of the recovered powder for the various pretreatments on the entire cell.

ID	Mean Area [µm2]	Mean Diameter [µm]
Cathode ref.	56.2	8.5
THCa-1	17.7	4.7
THCa-2	31.1	6.3
THCa-3	21.2	5.2
THCi-1	23.0	5.4
TEPC-1	51.6	8.1
TEPC-1R	49.8	8.0
TEPCC	49.4	7.9

. The particle size distributions of the recovered powder for the various pretreatments obtained from the image analysis of SEM results, used for the determination of the mean diameter, are displayed in Figure S10 of the Supplementary Materials.

The SEM analysis of THCa-1, shown in Figure 6a, reveals that the NMC particles are heavily damaged, with a mean diameter of approximately 4.7 µm. Furthermore, some fragments of the anodic copper have been found in the powder; this result is also confirmed by the EDS analysis (see Figures S11 and S12 in the Supplementary Materials). THCa-2, processed with a shorter time and lower stirring velocity, yields better results, with the aggregate structure appearing less damaged (Figure 6b). This confirms the hypothesis that the mechanical treatment of the particles can influence the final morphology of the recovered material [14]. Even though the mean particle diameter is increased to a value of 6.3 µm, this procedure remains unsuitable for industrial application. There are clear signs of damage and fragments of copper, as confirmed by the EDS analysis. The thermal treatment using sonication (THCa-3) gives the same results as THCa-2, with clear evidence of structural damage, and copper fragments are also detected in this case. For the thermal treatment in an inert atmosphere, the same conclusion can be made. In contrast, the use of TEP as a solvent for the recovery of the active material yields excellent results, with no evident structural damage and a mean particle diameter of 8.1 µm, which closely resembles that of the untreated cathode.

3.3. Pretreatment on a Cycled Battery

The pretreatment method using TEP was found the most promising among those tested and is therefore applied to the cycled battery cell. The result shows a good recovery efficiency of about 91% (see Figure 5), and the lost lithium in the recovery process is aligned with the previous experiments on a pristine cell. From the SEM images depicted in Figure 7, the recovered material appears intact with a mean particle diameter of 7.93 µm, which is lower compared with the performance on the pristine battery, possibly due to the degradation of the cathode during the cycles.

The cathode materials recovered from the most promising pretreatment procedures (THCa-1 and TEP for the pristine and cycled battery) were analyzed using XRD (Figure 8) and compared with a commercial NMC532 [36,37].

The material recovered using TEP in both cases exhibits peaks comparable to those of NMC532 (ICDD card number 01-086-2964). In the case of thermal pretreatment, however, two additional non-characteristic peaks, labeled A and B, are clearly visible. Lastly, the intensity of the peaks (003) and (101) is compared to each other. This parameter is crucial for assessing cationic mixing, which impacts the electrochemical properties of the material. When this ratio exceeds 1.2, the powder demonstrates favorable properties due to a low degree of cation mixing [37,38]. For THCa-1, the ratio between peaks 003 and 101 is 1.55, while, for TEPC-1, it is 1.49, and, for TEPCC, it is 1.45, confirming the efficacy of the pretreatment procedures. For the cycled battery, this ratio has a lower value, possibly due to the degradation that takes place during cycling.

4. Conclusions

In this work, we evaluated several pretreatment methods for recovering active cathode materials at two levels. Initially, the treatments were applied to cathode sheets to gauge their effects on recovery efficiency and material properties. In these tests, thermal, TEP, and NaOH pretreatments all yielded high efficiencies (>90%) with minimal alteration of the cathode stoichiometry. Notably, the TEP and thermal (TH-1 and TH-3) pretreatments preserved the pristine structure, maintaining a mean particle diameter of 8 µm. In contrast, NaOH pretreatment compromised the cathode’s integrity, resulting in a mean particle diameter of 6.3 µm or lower. For TH-1 and TH-3, lithium losses of approximately 10% were likely due to leaching into the solution. Subsequently, the most promising pretreatments were applied to whole cells. In this context, the presence of additional components, such as the anodic graphite, introduced further challenges. All the thermal pretreatments resulted in a mean particle diameter lower than 6.3 µm, indicating structural degradation. Conversely, TEP pretreatment preserved a nearly pristine structure with a mean particle diameter of 8 µm also for the cycled cell, with a recovery efficiency above 90%. Overall, these results highlight that the performance of pretreatment methods may differ when applied to previously separated cathode sheets or to entire cell configurations. This study underscores the potential of TEP as a green solvent for efficient cathode material recovery, given its high recovery efficiency, preserved stoichiometry, and minimal structural damage to the recovered powder.

Although TEP pretreatment demonstrates promising recovery efficiency and structural preservation, its cost and potential limitations with alternative battery chemistries require further exploration. Moreover, scaling this process for industrial operations necessitates a thorough economic and technical feasibility study. Future work should focus on optimizing process parameters, developing sustainable strategies for solvent recovery, and performing an economic analysis to assess the viability of the proposed approach.