Investigation of Recycling Behavior of Lithium Iron Phosphate Batteries with Different Thermal Pre-Treatments ^†

Abstract

Lithium iron phosphate (LFP) batteries are becoming a growing trend as a consequence of EU regulations and their advantages over nickel manganese cobalt (NMC) batteries. The use of LFP batteries is expected to increase considerably globally, creating an enormous waste problem. Battery recycling is emphasized in the EU’s battery laws, especially for lithium. Proper recycling of LFP batteries reduces environmental damage and supports the circular economy. Graphite and lithium are acknowledged as essential raw minerals. Recycling LFP batteries can be accomplished using pyrometallurgical and hydrometallurgical techniques. This study investigates the impact of thermal pre-treatments on flotation and leaching to produce an efficient recycling process.

1. Introduction

The escalating demand for lithium stems from its increasing use in sustainable technologies, including but not limited to mobile phones, laptops, electric cars, and power grids. The US Geological Survey’s mineral commodity report posits that approximately 20 million tons of lithium deposits are available for extraction. However, primary source production of lithium is a time-intensive and difficult process involving multiple stages, such as mining, roasting, acid baking, and evaporation, which demand significant energy consumption and chemical inputs. The production of secondary sources is a less complex and more efficient method compared to primary production. In particular, most of the global lithium mine production originates from a few regions, such as Western Australia, Chile, Bolivia, Argentina, and California. This results in the proper logistical arrangements of the extracted lithium to production plants, a costly process contributing to carbon emissions [1].

The first olivine-type lithium iron phosphate (LFP) batteries were synthesized in 1996. After the synthesis, the use of LFP batteries has grown in various modern applications, such as electric vehicles and large-scale energy storage devices, especially after the gasoline-powered vehicle restrictions by the EU. Since LFP batteries have lower cost, safer operation, high energy density, and a longer life cycle compared to nickel manganese cobalt (NMC) battery-type alternatives, it is expected to have increased demand similar to that in China. Globally, the use of LFP batteries grew from 12,500 tons in 2014 to 32,000 tons in 2015, with an expectancy of 64,000 tons in 2025 [2,3].

Naturally, this demand and use of LFP cells will create a huge waste of end-of-life products in the near future. Due to the components of the batteries, the waste could not be disposed of by nature. Moreover, the planned battery directive by the EU requires the recycling of batteries, with a special focus on lithium. Therefore, the correct recycling of LFP batteries will reduce environmental damage and contribute to the circular economy [3,4,5]. According to the European Commission’s list of critical raw materials (CRM) in 2023, lithium and graphite are listed as CRM for the future due to both their economic importance and supply risk [6].

LFP batteries can be recycled using both pyrometallurgical and hydrometallurgical methods. Processes start with discharging to avoid short circuits. Next, the cells may be dismantled or directly comminuted, depending on the recycling route. Thermal or chemical treatments are carried out before hydrometallurgical processing. Pre-treatment is normally not optional because at least the low-boiling organic has to be removed [1,7].

Hydrometallurgical processes may include leaching, solvent extraction, ion exchange, and precipitation methods for metal recovery, depending on the raw material type and final product purity. In the case of recycling LFP batteries, there are various methods. Li et al. investigated lower-concentration H₂SO₄ leaching in the presence of H₂O₂, followed by Li precipitation with Na₃PO₄ from the Li-containing leach solution. The Li leaching yield was 96.85%, and the precipitated yield of Li was 95.56% [3].

Zheng et al. crushed waste LFP, roasted it at 600 °C for 1 h, and then separated aluminum with a vibration sieve. The roasted and aluminum-separated sample was leached with a 2.5 mol/L H₂SO₄ acid solution at 60 °C and a liquid/solid ratio of 10 mL/g for 2 h. First, Fe precipitated as FePO₄ with pH adjustment. After removing Fe, Li precipitated with Na₂CO₃ [8].

LFP recycling can be accomplished through various methods, commonly involving leaching and precipitation. The leaching step aims to dissolve the elements of interest in the source material. Sulfuric acid (H₂SO₄) is commonly preferred as a leaching agent over hydrochloric and nitric acid due to its ease of handling, high yield, low cost, and less aggressive nature [2,9]. In some cases, the addition of oxidizing agents like hydrogen peroxide (H₂O₂) can be used in combination with sulfuric acid to increase the selectivity of lithium while removing iron from the solution by changing its valence [3].

In this study, different thermal pre-treatment temperatures were applied, and the effect of these thermal treatments on the following recycling operations was investigated with a focus on flotation and leaching. In addition, the removal of the binder and the organic solvents with thermal treatment was also investigated. The results from various temperature and thermal treatments are compared, and an optimized process route for a robust and highly efficient process is designed. A basic flowchart showing the differences between the investigated routes is given in Figure 1.

2. Experimental Procedure

2.1. Materials and Reagents

In this research, black mass sourced from a local German company was utilized.

The cathode material includes aluminum foil and binder. The anode materials include graphite, binder, and copper foil. Polyvinylidenfluorid (PVDF) is generally used as a cathode binder because of its good heat resistance and corrosion resistance.

The reagents used in this study—sulfuric acid (H₂SO₄), kerosene, and methyl isobutyl carbinol (MIBC)—were of analytical grade and purchased from Sigma-Aldrich (St. Louis, MO, USA).

2.2. Methods

The thermal pre-treatment experiments were conducted utilizing a Nabertherm muffle furnace devoid of atmosphere control, operating at varying temperatures. The experimental procedure commenced with the placement of the homogeneous sample into a nickel crucible. Following this, the crucible was positioned within the furnace at the predetermined temperature. Ultimately, the sample was extracted from the muffle furnace upon completion of the specified duration.

The LFP material was sieved using Haver Boeker analytical sieves with mesh sizes of 250 and 100 microns. The sieves were arranged as a sieving tower from top to bottom according to their mesh size. The original particle size was 211.82 microns, which is outside the particle size range suitable for flotation.

All flotation experiments transpired within the mechanical flotation cell type Denver. Initially, distinct thermally pre-treated samples underwent flotation under identical experimental conditions. Subsequent to this, solid–liquid separation was effectuated through vacuum filtration for both floated and depressed samples. Finally, all samples were subjected to drying at 105 °C for subsequent analytical assessment. In all flotation experiments, 12.5 g samples, 350 g/t of kerosene, and 150 g/t of MIBC were utilized, maintaining a fixed total volume of 125 mL.

The leaching experiments were conducted utilizing a three-neck flask. Initially, an acid solution was prepared, after which the temperature of the acid solution was stabilized at the specified temperatures. Following this, the black mass sample was gradually introduced into the solution. Once again, vacuum filtration was employed for solid–liquid separation, and both solid and liquid samples underwent subsequent analysis.

2.3. Characterization

The samples were analyzed regarding chemical composition and particle size distribution.

The particle size distribution of the sample was measured using laser diffraction with The Sympatec HELOS/KR system equipped with a RODOS dry dispenser.

The chemical composition of solid and liquid samples was analyzed using inductively coupled plasma optical emission spectroscopy (ICP-OES) with an Agilent 5100 ICP-OES system.

The determination of fluorine ion composition was measured using an ion chromatography method with an Ion chromatograph 761 Metrohm system. The Elemental analyzer EA4000 device was used for the determination of carbon from the initial and experimental samples.

3. Results and Discussion

3.1. Characterization of the LFP Black Mass

The LFP black mass composition is given in

Table 1. Chemical composition of the sample.

Element	Al	Fe	Li	P	F	C
Composition (%)	1.13	18.65	2.63	11.30	3.00	38.60

.

The particle size distribution of the sample is given in Figure 2. Approximately 80% of particles in the sample are below 154 μm, and 90% are below 215 μm. Particle size analysis results show that the initial sample is agglomerated.

3.2. Thermal Pre-Treatment

The optimal conditions for binder and electrolyte removal were investigated by analyzing the impact of thermal pre-treatment at various temperatures. The thermal pre-treatment was applied at 300, 400, and 500 °C. After thermal pre-treatment, weight loss and the chemical composition were calculated, and the results are given in

Table 2. Weight loss at different thermal pre-treatments.

Temperature (°C)	300	400	500
Weight loss (%)	4.36	5.96	6.75

.

The experimental results indicate that with an increase in thermal pre-treatment temperature, there is a corresponding increase in weight loss, reaching a certain point. However, it is important to note that higher-temperature experiments were not conducted, as the primary focus of this study was not to optimize the applied thermal treatment but to observe the effect of different temperatures on the following processing steps. The observed weight change can be attributed to the evaporation of a substantial portion of the electrolytes and binders.

Chemical analysis further revealed a reduction in the concentration of fluorine ions in the solution following thermal pre-treatment. The chemical analysis results for samples roasted at different temperatures are presented in

Table 3. Chemical composition of the untreated and treated sample.

Element	Al	Fe	Li	P	F	C
Without treatment (%)	1.13	18.65	2.63	11.30	3.00	38.60
300 °C (%)	1.08	18.06	2.69	11.41	2.83	38.04
400 °C (%)	1.09	18.23	2.61	11.24	2.45	38.05
500 °C (%)	0.97	18.15	2.80	12.13	2.08	35.58

. Notably, as the roasting temperature increases, the amount of fluoride ions in the sample decreases. This finding suggests that thermal pre-treatment also plays a role in effectively removing hazardous fluorine ions from the solution.

3.3. Flotation

In the first experiment, only a dried sample, which was not thermally treated, was employed. As expected, no separation between graphite and LFP was observed. This lack of separation can be linked to the inability to release the grains due to the presence of the binder, which was not removed with any thermal process. Subsequently, the flotation process was applied to the roasted samples at different temperatures. The flotation results are given in Figure 3, and the graphite’s purity is shown in Table 3.

Based on the results of the flotation experiments, it was observed that graphite follows the cathode and acts with Al, Cu, Fe, and Li when roasted at 300 and without roasting. At this temperature, there is no clear separation between graphite and Al, Cu, Fe, and Li due to the insufficient temperature for binder removal. However, at roasting temperatures of 400 and 500 degrees, 40% of the graphite was successfully removed via flotation. Notably, the graphite purity increased from 38% to 65% upon the application of flotation after proper thermal treatment. The purification results for graphite from the depressed and flotation samples are detailed in

Table 4. Purity of graphite (%).

	300 °C	400 °C	500 °C
Floated	65.13	66.54	59.10
Depressed	35.60	28.66	29.58

. It is evident that the floated material consistently exhibits higher purity compared to the depressed material at any given temperature. This can be explained by the fact that graphite has a natural hydrophobicity.

Additionally, the heat treatment at 500 °C was applied to both sieved and unsieved samples, and their flotation behaviors were examined. Sieving significantly contributed to the graphite purity, yielding 83.5% purity at the floated part compared to 59.1% purity reached without sieving.

This is a significant improvement in removing particles, such as aluminum and copper, present as foils in the black mass, thereby influencing the flotation behavior. The removal of such foil particles, particularly aluminum, is thought to enhance the effectiveness of the flotation process. The purities of the graphite from the depressed and flotation samples are detailed in

Table 5. Purity of graphite in 500 °C treated (%).

	500 °C (without Sieving)	500 °C (with Sieving)
Floated (C purity)	59.10	83.50
Depressed (C purity)	29.58	44.10

.

3.4. Leaching

The initial examination was performed with a black mass that was unsieved and unroasted to create a baseline for the comparison. Water and sulfuric acid were utilized and compared as lixiviants. The sulfuric acid concentration was calculated according to the reaction stoichiometry given in Equation (1).

2LiFePO₄ + 3H₂SO₄ → Li₂SO₄ + 2FeSO₄ + 2H₃PO₄

(1)

A stoichiometric amount of sulfuric acid, according to Equation (1), was preferred for the preliminary leaching tests. The results of the water and sulfuric acid experiments are presented in Figure 4. In both of these experiments, a constant solid/liquid ratio of 1/10, a leaching duration of 60 min, and a leaching temperature of 60 °C were used.

Water and sulfuric acid leaching results demonstrate that water leaching yields are relatively lower than sulfuric acid leaching under the given conditions. These results can be explained by water’s ability to dissolve only a small amount of lithium from electrolytes and a minimal portion of lithium present in the olivine LFP structure. In contrast, sulfuric acid actively breaks down the olivine LFP structure, facilitating the simultaneous dissolution of Li, Fe, and P. Hence, the subsequent studies focused on utilizing sulfuric acid as the lixiviant.

The effects of H₂SO₄ concentration and leaching duration were also investigated, which are presented in Figure 5. The conditions were kept constant (1/10 solid/liquid ratio and 60 °C) during the preliminary experiments.

According to the results of the experiments, even with a lower stoichiometry (0.8 STC), the acid concentration was enough to dissolve the Li at 40 min. After this point, the leaching yield was not effectively improved and remained stable. In all cases, more than 90% of the LFP cathode was found to have leached out of the black mass.

It is also found that applying different thermal treatments has an impact on the leaching behavior. Figure 6 shows the effect of thermal treatment temperatures on the leaching behavior.

As can be seen in Figure 6, an increase in leaching efficiency was observed compared to the ones without any thermal treatment. Almost complete dissolution was achieved when roasting temperatures of 300 °C and 400 °C were applied. Nevertheless, a lower leaching yield was observed when the thermal treatment reached 500 °C, possibly due to the formation of more stable oxide structures at higher temperatures.

4. Conclusions

In this study, the effect of thermal pre-treatment was investigated on leaching and flotation efficiency. The thermal pre-treatment directly affects the flotation behavior of graphite due to the removal of the binder and alteration of the surface properties. After removing the binder, it was found that the flotation of the graphite had improved significantly. Through the exclusive application of the flotation process, the purity of graphite, initially ranging between 35 and 38%, was elevated to 65%. However, this level of purity proves insufficient, necessitating further optimization to enhance the overall purity. Therefore, sieving was applied before flotation to separate not only the agglomerated particles but also the foil fraction from the source since the black mass was agglomerated as received. Both sieved and unsieved samples underwent a 500 °C pre-heat treatment, and their flotation behaviors were examined. The purity of the floated sample reached an impressive 84% for sieved samples, attributed to the removal of particles, especially aluminum, from the foil structure, thereby improving the flotation process’s effectiveness.

High leaching efficiencies were reached at 40 min at 60 °C, even with a lower stoichiometric ratio of the sulfuric acid. The thermal treatment also affected the leaching behavior positively, dissolving the cathode fraction almost completely. It is planned to focus on the kinetics of the leaching process, the optimization of the flotation for future investigations, and also the purification of Li from the leaching liquor.