High-Yield One-Pot Recovery and Characterization of Nanostructured Cobalt Oxalate from Spent Lithium-Ion Batteries and Successive Re-Synthesis of LiCoO₂

Abstract

A complete recycling process for the cathode material of spent lithium-ion batteries is demonstrated with a simple two-step process comprised of one-pot cobalt recovery to nanostructured materials and single step synthesis of LiCoO₂. For the facile and efficient recovery of cobalt, we employ malic acid as a leaching agent and oxalic acid as a precipitating agent, resulting in nanostructured cobalt oxalate. X-ray diffraction and Fourier transform infrared spectroscopy (FT-IR) analysis clearly show that cobalt species are simultaneously leached and precipitated as cobalt oxalate with a high yield of 99.28%, and this material can then be used as a reactant for the synthesis of LiCoO₂ for use as a cathode material. In addition to its advantages in simplifying the process, the proposed method allows for not only enhancing the efficiency of cobalt recovery, but also enabling reaction without a reducing agent, H₂O₂. Through successive single-step reaction of the obtained cobalt oxalate without any purification process, LiCoO₂ is also successfully synthesized. The effect of the annealing temperature during synthesis on the nanostructure and charge–discharge properties is also investigated. Half-cell tests with recycled LiCoO₂ exhibit a high discharge capacity (131 mA·h·g⁻¹) and 93% charge–discharge efficiency.

1. Introduction

Lithium-ion batteries (LIBs) have been considered the best technology for sustainable transport and smart electronics due to their excellent properties, including high energy density, long storage life, low self-discharge efficiency, and a wide operating temperature [1,2,3]. They are now extensively used as power sources in various portable electronic devices (e.g., smart phones), and their potential applications are also expanding to electric vehicles (EVs) and hybrid EVs (HEVs) [4,5]. However, LIBs in most portable devices have a lifespan of less than 3 years, and those in HEVs and EVs are projected to have a lifespan of roughly 10 years [4,6]. Given these limited lifespans and increasing production, it must be noted that equal amounts of spent LIBs are produced after they have reached the end of their useful life. Although spent LIBs are generally not classified as hazardous waste, their inappropriate disposal may cause environmental pollution because of the presence of toxic compounds. From an economic perspective, high-value nonferrous metals such as cobalt and nickel are contained in the positive electrode (cathode) [7,8,9]. In the cathode of spent LIBs, the concentration of metallic ingredients has been reported to be in the following order: cobalt (5–20%) > nickel (5–10%) > lithium (5–7%) [10], because the available cathode materials are mostly used in a form of LiCoO₂ and some alternatives such as LiNi_0.33Mn_0.33Co_0.33O₂, LiNi_0.8Co_0.15Al_0.05O₂, and LiFePO₄. Among these components, the recovery of cobalt is the most economically feasible since cobalt is about twice as expensive as nickel and 15 times more expensive than copper [11]. Furthermore, cobalt is classified as carcinogenic, mutagenic, and toxic to human health. Therefore, the recovery of cobalt from spent LIBs is required in order to prevent environmental pollution from spent LIBs as well as keep pace with the expansion of LIBs’ use in various applications [9].

Until now, several approaches have been developed to recover the cobalt species from spent LIBs, including pyrometallurgical [12,13], biometallurgical [14,15], electrochemical [16,17,18], and hydrometallurgical methods [10,19,20,21,22]. Among them, hydrometallurgical processes are most widely used for the complete recovery of metals with high purity and low energy consumption. They mainly involve acid leaching, chemical precipitation, separation, and electrochemical recovery [10]. In the leaching step, strong inorganic acids such as H₂SO₄, HCl, and HNO₃ have conventionally been used as leaching agents, with a high extraction efficiency of cobalt. Inorganic leaching agents release toxic gases like SO₃, Cl₂, and NO_x, and the waste acidic solution causes serious secondary damage to the environment and human health. Moreover, the recovery processes are still complicated due to the multiple steps required, including the separation and purification processes of cobalt and other components.

Thus, various organic acids have been extensively studied as environmentally benign leaching agents due to their favorable properties, including ease of degradation, recyclability, absence of toxic gases, and the avoidance of secondary pollution. For example, Li et al. reported that citric acid, ascorbic acid, and malic acid could be successfully used for cobalt recovery from the cathode materials of spent LIBs [19,21,23]. However, the leaching efficiencies of the organic acids were significantly lower than those of sulfuric acid, even with the addition of sufficient hydrogen peroxide, which leads to a more facile separation and purification process for commercialization [24]. Therefore, a simple and environmentally acceptable process is urgently required to avoid adverse environmental impact and the mechanical complexity of the battery recycling process.

In this study, a novel one-pot recovery process was designed using malic acid and oxalic acid as a leaching agent and a precipitator, respectively, which is more efficient than the previous process wherein the leaching process was separated from precipitation. The cobalt species were efficiently leached and simultaneously precipitated by an aqueous mixture of malic acid and oxalic acid in a single process. It was also found that our process could be successfully operated without the assistance of a reducing agent via Le Chatelier’s principle. In addition, the precipitated cobalt oxalate compound was used to prepare the cathode material of LIBs, LiCoO₂, in a single-step reaction without purification process.

Therefore, our novel process successfully demonstrated a complete recycling process of the cathode material of spent LIBs in a two-step process composed of one-pot cobalt recovery and single-step synthesis of LiCoO₂. This is considered to simplify the Co recovery process compared to previous recovery processes of spent LIBs that are composed of multiple steps such as leaching, chemical precipitation, solvent extraction, and synthesis of cathode material.

2. Experimental

2.1. Screening of Organic Acids

The cathode material was separated from spent LIBs for cellphones, and was kindly supplied by Sungeel HiTech Co. Ltd (Incheon, Korea). Before the leaching experiments, the cathode material was thermally calcined at 700 °C to remove organic impurities. For the selection of an environmentally-benign leaching agent, various organic acids were used without further purification.

Leaching experiments were conducted in a 0.5 L three-necked round-bottomed thermostatic Pyrex reactor equipped with a temperature control facility, reflux condenser, and a mechanical impeller. A measured amount of the cathode material was added to a 2 M aqueous acid and 5 wt % H₂O₂ solution with a solid/liquid ratio of 1/10. The system was maintained at 80 °C under constant magnetic agitation of 300 rpm for 3 h. Leach liquor samples of 5 mL were periodically drawn at desired time intervals. After filtering the samples, the metal ion concentration in the solution was analyzed by inductively coupled plasma atomic emission spectroscopy.

2.2. One-Pot Process

An aqueous mixture of malic acid and oxalic acid was used as leaching and precipitation agent, respectively, for the demonstration of simultaneous leaching and precipitation in a single reactor. The total concentration of acids was fixed at 2 M, and the molar ratio of malic acid to oxalic acid was 1 to 1. The precipitated cobalt oxalate was filtered and washed with distilled water. The phase purity and chemical structure of cobalt oxalate was characterized by X-ray diffraction (XRD, Rigaku, Tokyo, Japan) and Fourier transform infrared spectroscopic (FT-IR, Shimadzu, Kyoto, Japan) analysis.

2.3. Synthesis of LiCoO₂ Powder

To explore a one-step process for the synthesis of LiCoO₂ powder, the cobalt oxalate obtained from the one-pot process was mixed with a lithium precursor (Li₂CO₃ or LiOH) in a mortar with a Li/Co ratio of 1.05/1. Subsequently, the mixtures were calcined in a muffle furnace at different temperatures from 400 °C to 900 °C for 8 h. The products were thoroughly washed with distilled water and dried at 80 °C for 24 h. The morphology of the synthesized LiCoO₂ powder was characterized by field emission scanning electron microscopy (FE-SEM, Thermo Fisher Scientific, Waltham, MA, USA) and XRD.

2.4. Electrochemical Measurements

Cyclic voltammetry and galvanostatic charge–discharge experiments were conducted on individual electrodes in a typical three-electrode cell with 1 M Li₂SO₄ aqueous solution. A platinum wire and a saturated calomel electrode (SCE) were used as the counter and reference electrodes, respectively. A Luggin capillary faced the working electrode at a distance of 5 mm. The working electrodes were prepared by mixing active material, acetylene black, and poly (tetrafluoroethylene) (PTFE) in a weight ratio of 85:10:5, and pressing the powdered mixture onto a titanium mesh.

3. Results and Discussion

3.1. Screening Test of Organic Acids

A systematic screening procedure for various organic acids was performed based on leaching efficiency. Considering the chemical structure and reaction constant, eight candidate leaching agents from malic acid to acrylic acid were tested, as shown in Figure 1. The results showed that sulfuric acid, citric acid, and malic acid exhibited higher leaching efficiency compared with the other six organic acids, at 95.8%, 89.9%, and 86.69%, respectively. The carboxyl and hydroxyl groups in citric and malic acid are expected to give rise to H⁺ in distilled water [23,24]; therefore, dissociated carboxyl groups will react with cobalt ions upon the addition of the reducing agent H₂O₂, the ability of which to enhance the aqueous solubility of cobalt ions is also well known [22,25]. Among these organic acids, malic acid was chosen in this study for further investigation of a one-pot leaching and precipitation process to re-synthesize LiCoO₂, thanks to its advantages in price and accessibility over citric acid.

A schematic view of the one-pot process proposed here is given in Figure 2. With the addition of oxalic acid into a solution with malic acid, cobalt ions are expected to simultaneously undergo a leaching and precipitation process, resulting in cobalt oxalate. Oxalic acid is a common weak acid, and it is well-known that its anion acts as an excellent ligand to easily form coordination compounds with metal ions thanks to its dibasic characteristics [10,26]. Figure 3 shows the optical image, scanning electron microscopy (SEM) image, and XRD pattern of the precipitated powder from the one-pot process. They clearly indicate that cobalt oxalate forms during the one-pot process and grows to a rod-like shape with a (001) direction that is attributed to hydrogen bonding during the reaction, as previously reported [27,28]. FT-IR analysis also shows several peaks corresponding to hydrated metal oxalates (Figure 4). The broad band at 3360 cm⁻¹ is attributed to the water of hydration. The band at 1622 cm⁻¹ corresponding to asymmetric ν(C–O) and the closely spaced bands at 1362 cm⁻¹ and 1317 cm⁻¹ assigned to symmetric ν(C–O), respectively, clearly reveal the presence of bridging oxalates, with all four oxygen atoms coordinated to the cobalt atoms.

Furthermore, oxalic acid allows for a process that does not employ H₂O₂ as a reduction agent, without any loss of recovery efficiency (Figure 5). Previously, hydrometallurgical processes for recycling cobalt from spent lithium-ion battery have included the addition of H₂O₂. The presence of H₂O₂ significantly favors the leaching of cobalt, since H₂O₂ can reduce the Co³⁺ to Co²⁺ which is more readily dissolved than Co³⁺ in malic acid (C₄H₆O₅) solution as described in Equation (1). In the absence of H₂O₂, the leaching efficiency of cobalt reached only 40%, as shown in Figure 5. In our one-pot process, however, the released Co³⁺ from the spent cathode materials immediately react with oxalic acid (C₂H₂O₄) to form a cobalt oxalate precipitate; in other words, aqueous cobalt ions are removed from the solution (Equation (2)). Hence, the removal of Co³⁺ ions shifts the chemical equilibrium to facilitate the dissolution of Co³⁺ ions by Le Chatelier’s principle, as shown in Figure 2.

2LiCoO₂ (s) + 3C₄H₆O₅ (aq.) + H₂O₂ → 2Co²⁺ (aq.) + 2Li⁺ + 3C₄H₄O₅²⁻ (aq.) + 4H₂O + O₂ (g)

(1)

2LiCoO₂ (s) + C₄H₆O₅ (aq.) + 2C₂H₂O₄ (aq.) → 2CoC₂O₄ (s)↓ + 2Li⁺ + C₄H₄O₅²⁻ (aq.) + 2H₂O

(2)

Consequently, our one-pot process eliminates the need for the reduction agent H₂O₂ to reduce the poorly soluble Co³⁺ to the highly soluble Co²⁺. The cobalt recovery efficiency in our process is enhanced to 99.28% compared with the processes using only oxalic acid or malic acid/H₂O₂, which have recovery efficiencies of 44.10% and 86.69%, respectively. Therefore, an efficient one-pot process to acquire Co species was achieved, whereas the leaching and precipitation processes are separated in the other conventional method.

3.2. Synthesis of LiCoO₂ Powder from Co-Oxalate

To re-synthesize LiCoO₂ from the obtained Co-oxalate, LiOH and Li₂CO₃ were added as lithium precursors, and the mixture was calcined from 400 °C to 900 °C. XRD shows that LiCoO₂ is synthesized with the crystal structure of the layered α-NaFeO₂-type structure in the R3m space group from Co-oxalate after calcination (Figure 6) [29,30]. The calcination process is known to induce thermal decomposition as well as dehydration of Co-oxalate, thus removing carbon and other residual species [27,28]. Note that while the peak corresponding to Co₃O₄ also appears in the sample from reaction with Li₂CO₃, only the crystalline planes of the LiCoO₂ peaks are indexed after reaction with LiOH. It is expected that the high melting point of Li₂CO₃ and oxygen-rich environment facilitates the oxidation of Co, hence producing Co₃O₄. Other hydrometallurgical recovery processes resulting in CoSO₄ or Co-oxalate require a two-step process to produce LiCoO₂, involving the oxidation of cobalt sulfate and lithiation of cobalt oxide [20]. In contrast, our process is beneficial, as it produces LiCoO₂ using a single-step reaction with LiOH. This result demonstrates that LiCoO₂ can be directly synthesized from the reaction of the precipitated Co-oxalate with LiOH without generating the by-product Co₃O₄, which would require an additional process to refine. To the best of our knowledge, this is the first demonstration of a single-step synthesis of LiCoO₂ from Co-oxalate that is recycled by a hydrometallurgical process. We further investigate the effect of calcination temperature on the microstructure and morphology of the re-synthesized LiCoO₂. SEM was used to observe the change in morphology of LiCoO₂ with temperature. As shown in the SEM image (Figure 7a), the rod shape of LiCoO₂ changed to a particular shape with increasing calcination temperature, suggesting dehydration and thermal decomposition of Co-oxalate during calcination. Furthermore, the resulting product shows a finer particle-like structure with increasing temperature due to densification of LiCoO₂ in calcination. XRD results also support the assertion that the crystallinity of the re-synthesized LiCoO₂ is enhanced with annealing temperature (Figure 7b). With increasing calcination temperature, the peak intensity also increases, and high-order peaks such as (009) and (105) appear, which are attributed to the enhanced crystallinity of LiCoO₂. Note that the (104) peak intensity dramatically increases with temperature, indicating high crystallinity in the c-axis direction. The calculated degree of cation mixing with the ratio of the peak intensity (003) to (104) was 2.17 at 900 °C, 2.04 at 700 °C, and 1.85 at 400 °C [31,32,33]. It could be expected that LiCoO₂ calcined at a high temperature of 900 °C will have good electrochemical properties, which will be discussed in the next section. The grain size analyzed based on the full width at half maximum (FWHM) of the (003) peak increased from 6.1 nm at 400 °C to 13.7 nm at 900° C, which is in a good agreement with the enhanced crystallinity with high temperature calcination. We believe that calcination at high temperature induces growth along the (001) plane (which has the lowest surface energy) in order to minimize the surface energy, hence resulting in high crystallinity as well as a high degree of cation ordering [34,35].

3.3. Electrochemical Characterization

The electrochemical properties of re-synthesized LiCoO₂ calcined at high temperature were characterized using a cyclic voltammetry test. Figure 8 clearly shows that the peak for Li insertion occurs at 0.77 V and a desertion peak occurs at 0.53 V, indicating that the electrochemical reaction occurs successfully in LiCoO₂. However, in LiCoO₂ calcined at 400 °C, the Li-ion insertion peak appears at 0.51 V, corresponding to the electrochemical reaction of arbitrary cobalt oxide, which can be referred to as CoO_x, suggesting that calcination at low temperature can generate by-product cobalt oxide residue, as previously reported [28]. The discharge capacity was also measured by a half-cell test for samples calcined at 400 °C, 700 °C, and 900 °C at a rate of 0.5 C (Figure 9). The voltage profiles show that the discharge capacity was improved from 79.8 mA·h·g⁻¹ in the sample calcined at 400 °C to 131 mA·h·g⁻¹ in the one at 900 °C with increasing calcination temperature. This result is in good agreement with our XRD results, indicating that high temperature calcination induces a high degree of cation ordering [31,36]. It has been reported that the Li ion inserts more efficiently for a sample with a lower degree of cation ordering in a textured electrode structure thanks to a greater number of electrochemical active sites and a shorter diffusion length than a sample with a higher degree of cation ordering [30,35,37]. However, our electrodes were processed using powder in this study, and hence no preferential diffusion path of Li ions was introduced, in contrast to previous textured thin-film-based electrodes [38,39]. Furthermore, the re-synthesized LiCoO₂ materials show very low values for the degree of cation ordering (~2), suggesting a short diffusion channel similar to a sample previously reported with a degree of cation ordering greater than four Therefore, it is expected that high impedance resulting from a high degree of crystallinity will have a more dominant effect on the electrochemical characteristics than the Li ion diffusion path [33,40].

Figure 9b compares the rate capability up to a 10 C rate for the samples calcined at 400 °C, 700 °C, and 900 °C. The LiCoO₂ calcined at 900 °C exhibits a rate capability of 100 mA·h·g⁻¹ at a rate of 1 C and 82 mA·h·g⁻¹ at a rate of 10 C. The discharge capacity of re-synthesized LiCoO₂ initially decreases over a few cycles, which is expected due to some residue on the surface of LiCoO₂ generated during our process, reacting with the electrolyte and resulting in a surface interface that increases the electrode impedance and induces poor power performance [41,42]. Improvement of the capacity retention via surface analysis and thorough element control will be further investigated in the near future.

4. Conclusions

In summary, a novel process to produce LiCoO₂ from spent batteries has been developed using a one-pot process including both leaching and precipitation of Co to re-synthesize LiCoO₂ through reaction with LiOH. Co-oxalate can be used for re-synthesis of LiCoO₂, and is obtained from spent batteries using malic acid as a leachant and oxalic acid as a precipitator. This process is shown to yield cobalt species with a high efficiency of 99.28% and achieve a process without the assistance of H₂O₂. We also find that an additional single-step reaction of Co-oxalate with LiOH produces high-quality LiCoO₂ without producing by-product cobalt oxide species; hence, an additional refinement process is not required. Our XRD and electrochemical characterization also suggest that the resulting LiCoO₂ shows high crystallinity and good electrochemical properties as a cathode in LIBs, similar to the commonly-used LiCoO₂. As a result, our novel process successfully demonstrated a complete recycling process of the cathode material of spent LIBs in a simple two-step process, which is composed of one-pot cobalt recovery and single step synthesis of LiCoO₂. In addition, our recovery process is expected to potentially be environmentally friendly due to the removal of H₂O₂ in the recovery process. The environmental assessment of our process will be further investigated in a following report.