Open AccessArticle

Mechanism and Experimental Study on the Recovery of Rare Earth Elements from Neodymium Iron Boron Waste Using the ZnF₂ Fluorination Method

School of Metallurgical Engineering, Jiangxi University of Science and Technology, Ganzhou 341000, China

Yichun Lithium New Energy Industry Research Institute, Jiangxi University of Science and Technology, Yichun 336000, China

Fengcheng Nonferrous Metals and Advanced Materials Industrial Research Institute, Jiangxi University of Science and Technology, Fengcheng 331100, China

⁴

Jiangxi Chunhua Lithium Industry Co., Ltd., Fengcheng 331100, China

Authors to whom correspondence should be addressed.

Materials 2024, 17(23), 5807; https://doi.org/10.3390/ma17235807

Submission received: 9 September 2024 / Revised: 8 November 2024 / Accepted: 17 November 2024 / Published: 27 November 2024

Download

Browse Figures

Figure 1
XRD pattern of NdFeB waste. "> Figure 2
Plot of the reaction <math display="inline"><semantics> <mrow> <msubsup> <mrow> <mi mathvariant="sans-serif">Δ</mi> <mi>G</mi> </mrow> <mrow> <mi>T</mi> </mrow> <mrow> <mi>θ</mi> </mrow> </msubsup> </mrow> </semantics></math> versus T for the roasting process. "> Figure 3
(a) Effect of the ZnF2-to-raw material mass ratio on rare earth element fluorination rate; (b) XRD spectra of the clinker with different ZnF2 dosages. "> Figure 4
(a) Effect of roasting temperature on rare earth element fluorination rate; (b) XRD spectra of the clinker with different roasting temperatures. "> Figure 5
(a) Effect of roasting time on leaching; (b) XRD spectra of the clinker with different roasting times. "> Figure 6
Comparison between predicted and actual values of rare earth element fluorination rates. "> Figure 7
Surface plot of the response of the interaction between different factors on the rare earth element fluorination rate: (a) roasting temperature and roasting time, (b) roasting temperature and mass ratio of ZnF2 to raw material, and (c) mass ratio of ZnF2 to raw material and roasting time. "> Figure 8
XRD graph of leached products. "> Figure 9
Scanning electron microscope images of the leachate product at different scales: (a) 1 µm; (b) 500 nm; (c) 200 nm; (d) 100 nm. "> Figure 10
Process flow for recovering rare earths from NdFeB waste through the fluorination reaction. ">

Versions Notes

Abstract

We conducted a mechanistic and experimental study on zinc fluoride roasting for the recovery of NdFeB waste to address the difficulties faced during this pyrometallurgical recovery process, such as the high dependence on the quality of raw materials, the high energy consumption involved in roasting transformations, and the low added value of mixed rare earth products. Thermodynamic calculations showed the feasibility of fluorinating rare earths in NdFeB waste, and one-factor experiments were performed. The results showed that at a roasting temperature of 850 °C, a reaction time of 90 min, and 100% ZnF₂ addition, the fluorination rate of rare earths could reach 95.69%. In addition, after analyzing the mesophase composition of a clinker under different roasting temperature conditions, it was found that, when the roasting temperature exceeded 850 °C, the fluorination rate of rare earths was reduced, which was consistent with the thermodynamic results. On this basis, response surface methodology (RSM) was used to carry out experiments to investigate in depth the effects of various factors and their interactions on the fluorination rate of rare earths, which provides a sufficient experimental basis for the recovery of NdFeB waste via fluorination roasting. The results of this study show that ZnF₂ addition had the greatest influence on the rare earth fluorination reaction, followed by roasting temperature and roasting time. According to the optimization results of the model, the optimal roasting conditions were determined as follows: 119% ZnF₂ addition at 828 °C, a roasting time of 91 min, and a rare earth element fluorination rate of 97.29%. The purity of the mixed fluorinated rare earths was as high as 98.92% after leaching the roasted clinker with 9 M hydrochloric acid at a leaching temperature of 80 °C, a liquid–solid ratio of 4 mL/g, and a leaching time of 2.5 h. This study will lay the foundation for promoting the application of pyrometallurgical technology in the recycling of NdFeB waste.

Keywords:

NdFeB scrap; fluorinated reaction; Box–Behnken design; mixed fluorinated rare earths

1. Introduction

NdFeB magnets are a new type of rare earth functional material with excellent comprehensive performance, low cost, and easy processing, which have been widely used in wind power, hybrid vehicles, and hydraulic turbine generators [1,2,3,4]. Cutting and grinding processes must be carried out during the preparation of NdFeB magnets, and the amount of waste generated during these processes amounts to roughly 25%. By 2030, the market demand for NdFeB magnets will reach a minimum of 240 kilotons and a maximum of 644 kilotons. Concurrently, the amount of waste NdFeB magnets supplied for secondary recycling will reach a minimum of 27 kilotons and a maximum of 54 kilotons. It is evident that the recycling of NdFeB magnets has huge economic value [5,6,7,8]. Since NdFeB magnets are important materials and require a relatively high amount of rare earth elements, recovering these elements from NdFeB magnet waste has significant economic and environmental benefits [9].

Pyrometallurgical processes are high volume and quick, and different rare earth products (rare earth alloys or rare earth metals) can be obtained by choosing different processes, leading to broad application prospects [10,11]. The currently reported pyrometallurgical processes for NdFeB recycling mainly include selective chlorination, liquid–metal extraction, and fluoridation separation [12,13,14,15]. The selective chlorination method for recycling used magnets is based on the fact that rare earth elements are more reactive and combine more readily with chloride ions to form chlorides [16,17]. Lorenz and Bertau [18] investigated the treatment of NdFeB magnets with NH₄Cl, followed by leaching with acetic acid. Under optimal conditions, the yield reached only 84%. The liquid–metal extraction method uses the difference in the affinity of rare earth elements (REs) and iron for other metals to effectively enrich and separate rare earth elements and iron [19]. In particular, the rare earth element, Nd, can form low-melting-point alloys with Ag, Mg, etc. Sun et al. [20] investigated the behavior of Mg in Nd extraction from industrial NdFeB magnet waste. To prevent the oxidation of NdFeB magnets, the optimum extraction temperature should be below 830 °C. Liquid–metal extraction is not suitable for oxidizing waste or sludge material. The presence of rare earth oxides prevents the dissolution of rare earths in molten magnesium. The fluoride separation method utilizes the fact that the rare earth elements in NdFeB waste are more attractive to fluorine than to iron. Sung Gue Heo et al. [21] compared the effectiveness of fluoride and chloride by carrying out an exchange reaction of magnesium halides (fluoride and chloride) with NdFeB waste. The extraction rate of Nd was 98.64% for the fluoride system, at a roasting temperature of 873 K and a holding time of 12 h, and 84.59% for the chloride system, at a roasting temperature of 1073 K and a holding time of 12 h. Although magnesium fluoride has a high extraction rate for rare earths, the higher holding time increases energy consumption. Hu et al. [22] reported a method for the recovery of rare earth elements from scrap NdFeB using an AlF₃-NaF melt as the fluorinating agent, and the fluorinated rare earths obtained could be subsequently treated with molten salt electrolysis to directly produce Al-Nd-based alloys; however, this method requires fluorination at 900 °C for 3 h, and the recovery rate of rare earths is only 83%.

There are some areas in which current pyrometallurgical processes require urgent improvement, including (1) strict requirements regarding the quality of the waste material (the scrap must have a low content of impurities, and it needs to be tested and sorted to ensure that the content and proportion of the main components meet the requirements for recycling), with most methods applying to NdFeB alloy waste only, and (2) the low purity of the mixed rare earth elements resulting in a product that cannot be easily utilized. The selective fluorination process is considered one of the pyrometallurgical processes able to overcome the above technical difficulties [23]. Its ability to overcome these difficulties can be explained as follows: (1) rare earth elements are easier to combine with fluorine, and therefore, it is not necessary to apply strict requirements regarding the type and quality of the waste utilized during the fluorination reaction process, and (2) most recovered rare earth products require electrolysis to be converted into metals, while the direct production of fluorinated rare earths increases their added value [21,24]. However, limited studies on fluoride roasting for NdFeB scrap recycling result in insufficient theoretical and process parameter support. Therefore, there is an urgent need for an in-depth study of fluoride roasting for NdFeB scrap recycling to promote the application of pyrometallurgical processes in this field.

ZnF₂ is stable and does not release harmful gasses at high temperatures. We chose ZnF₂ as the fluorinating agent in the one-factor condition experiment to study the NdFeB fluoridation recovery’s reaction mechanism and experimental rule. Based on the results of our one-factor experiments, a response surface experiment was designed to establish a mathematical model of the fluorination process of rare earths according to the RSM criterion. The roasting process was optimized by analyzing the experimental results. On this basis, a mixed fluorinated rare earth product with a purity of 98.92% was obtained by acid leaching. Due to their high purity, fluorinated rare earths can be roasted at high temperatures to produce rare earth fluorine oxides, which can be dissolved in acid to obtain high-purity rare earth materials for further separation or directly utilized in the production of rare earth alloys or mixed rare earth metals through molten salt electrolysis. The research results will lay the foundation for promoting the application of pyrometallurgical technology in the recycling of NdFeB waste.

2. Materials and Methods

2.1. Materials

The raw material used in this study was provided by a NdFeB waste recycling factory in Ganzhou, China, including waste NdFeB magnets and waste materials generated from production. Firstly, the raw material was crushed and roasted to obtain NdFeB oxidized waste. An ICAP-PRO (full-spectrum, direct-reading, inductive coupled plasma emission spectrometer) determined the rare earth content in NdFeB oxidized waste. The proportions of various rare earth oxides are shown in Table 1. The content of rare earth oxides was 26.98%, whereby Nd₂O₃, CeO₂, Pr₆O₁₁, and Gd₂O₃ accounted for 94.82% of the rare earth elements. In addition, other impurity elements in the waste were analyzed using an ICAP-PRO analyzer and the chemical titration method, and the results are shown in Table 2. Besides rare earths, there was 48.22% of iron and other trace elements. From the XRD analysis results in Figure 1, it can be seen that the main components of the waste were Fe₂O₃ and rare earth oxides.

2.2. Procedure

The NdFeB scrap was ground for 6 min using a vibrating mill at 900 rpm to obtain 200-mesh raw materials, which were then mixed with ZnF₂ (50–150% addition) and subsequently continuously roasted in a muffle furnace at a constant roasting temperature of 750–950 °C for 30–150 min. After roasting, the roasted clinker was ground in a vibrating mill at 900 rpm for 1 min and then placed in a three-neck flask. The roasted clinker was stirred with 5% ammonia at a liquid–solid ratio of 50 mL/g for 60 min at 150 rpm at room temperature and then filtered to remove unreacted ZnF₂. The filter residue was dried at 80 °C for 12 h and placed in a three-necked flask and acid leached in a constant-temperature water bath at 80 °C for 2.5 h with the addition of 9 M hydrochloric acid at a liquid–solid ratio of 4 (mL/g) with stirring at 150 rpm, then the acid leachate was filtered. The acid leach residue was dried at 80 °C for 12 h. Next, 0.1000 g of the acid leach residue was dissolved in aqua regia and then analyzed for the purity of the mixed fluorinated rare earths using an ICAP-PRO instrument. The leachate was diluted 50 times with 5 % nitric acid and analyzed for rare earth concentration using an ICAP-PRO instrument. The leaching rate of each REE can be calculated by Equation (1); the total rare earth leaching rate can be calculated by Equation (2), and the total rare earth fluorination rate can be calculated by Equation (3).

η_{i} = \frac{C_{i} \times V}{m_{0} \times ω_{i}} \times 100 %

(1)

η_{i} = \frac{C \times V}{m_{0} \times ω} \times 100 %

(2)

χ = 1 - η

(3)

where

m_{0}

is the mass of the raw material;

ω_{i}

(%) is the content of rare earth element “i” in the raw material;

C_{i}

(g/L) is the concentration of rare earth element “i” in the leaching solution; and V (L) is the volume of the leaching solution.

ω

(%) is the raw material’s total content of rare earth elements; C (g/L) is the total concentration of rare earth elements in the leaching solution.

In this study, the raw materials were roasted in a KSL-1200X muffle furnace manufactured by Hefei Kejing Materials Technology Co. (Hefei, China). The rare earth content in the raw material, the concentration of rare earths in the leach solution, and the content of fluorinated rare earths in the leach residue were measured using an ICAP-PRO instrument from Thermo Fisher Scientific (Waltham, MA, USA). A field-emission scanning electron microscope (SEM), model MLA650F, produced by FEI (Hillsboro, OR, USA), was used to analyze the microscopic morphology of the leaching slag. The physical composition of the raw materials, roasted sand, and acid leaching slag was analyzed using an Empyrean-type X-ray diffractometer manufactured by Panalytical, (Almelo, The Netherlands). The main test specifications and parameters of the instrument were as follows: the target material was Cu-Kα (λ = 0.15406 nm), which was tested in a continuous scanning mode (2θ = 10°–90° in steps of 0.013°). The Gibbs free energies of the chemical reactions were calculated using the HSC 6.0 software developed by Outotec (Helsinki, Finland), and the temperature was set to 0–1000 °C.

2.3. RSM Optimization Process Based on Box–Behnken Design

The effects of roasting temperature, roasting time, and ZnF₂ dosage on the fluorination rate of rare earths were investigated using the RSM method. Three different levels were selected based on the results of the one-factor test. The Box–Behnken design (BBD) was implemented for the three levels of the three factors. Table 3 shows the codes designed and the corresponding actual levels according to the Box–Behnken design model in the Design Expert 13 Trial. Based on the results of the 17 group design experiments, analysis of variance (ANOVA) was performed using RSM, and the quadratic polynomial shown in Equation (4) was obtained by fitting [25].

y = a_{0} + \sum_{i = 1}^{3} a_{i} x_{i} + \sum_{i = 1}^{3} a_{i i} x_{i}^{2} + \sum_{i = 1}^{2} \sum_{j = i + 1}^{3} a_{i j} x_{i} x_{j} + ε

(4)

where y is the response value of RSM (rare earth fluorination rate);

a_{0}

is the intercept term obtained via ANOVA regression using least squares;

a_{i}

is the first-order linear coefficient;

a_{i i}

is the square effect;

a_{i j}

is the quadratic effect between factors;

x_{i}

and

x_{j}

are uncorrelated factors; and

ε

is the random error term caused by inconsistency between the model predicted and actual measured values.

3. Results and Discussion

3.1. Thermodynamic Calculation

Thermodynamic calculations are essential to analyze the reaction process under study [26]. From the second law of thermodynamics, it is known that, as the absolute value of

{Δ G}_{T}^{θ}

increases, the tendency for the chemical reaction to proceed is greater [27]. Using HSC6.0 thermodynamic software, the positive and negative values of the

{Δ G}_{T}^{θ}

value of each chemical reaction were calculated and compared separately and used as a basis to determine the chemical reaction order and reaction trend. The main chemical reactions of rare earth elements with ZnF₂ are listed in Table 4, and the calculated

{Δ G}_{T}^{θ}

values are shown in Figure 2.

From Figure 2, it can be seen that the

{Δ G}_{T}^{θ}

of reactions 1 to 6 tends to increase with the increase in reaction temperature, in comparison, that of reaction 7 tends to decrease with the increase in reaction temperature, which indicates that the increase in temperature is not conducive to the fluorination of Nd, Ce, Gd, La, Ho, and Dy but that it is conducive to the fluorination of Pr. For reactions 1 to 7,

{Δ G}_{T}^{θ}

is always negative, indicating that rare earth elements can react with ZnF₂ to form fluorinated rare earths; whereas for reaction 8,

{Δ G}_{T}^{θ}

is always positive, indicating that Fe₂O₃ cannot react with ZnF₂.

3.2. Single-Factor Condition Experiments

3.2.1. Effect of ZnF₂ Addition on the Fluorination Rate of Rare Earths

In order to study the effect of different ZnF₂ additions on the fluorination rate of rare earth elements, the conditions of the roasting process were set as follows: the roasting temperature was 850 °C, and the roasting time was 90 min. Figure 3a illustrates the impact of the leaching rate of rare earth elements at 50–150% of ZnF₂ addition (in Figure 3a, Figure 4a, and Figure 5a, REF denotes fluorinated rare earth elements, while REEs refers to total rare earth elements). From the curves of the different ZnF₂ additions shown in Figure 3a, it can be seen that the fluorination rate of rare earths increased from 81.21% to 95.69% when the ZnF₂ addition increased from 50% to 100%, but the fluorination rate of the rare earths slightly decreased when the ZnF₂ addition increased to 125%. As shown in Figure 3b, the diffraction peaks of the fluorinated rare earths were gradually enhanced when the addition of ZnF₂ increased from 50% to 100%, but the diffraction peaks of fluorinated rare earths decreased slightly when the addition of ZnF₂ increased to 125%. Kinetically, increasing the ZnF₂ content increases the chance of collision between the elements, which is favorable to the fluorination reaction. Therefore, when the addition of ZnF₂ increased from 50% to 100%, the fluorination rate also increased. However, when the amount of ZnF₂ was 125%, the fluorination rate of the rare earths decreased slightly, indicating that the reaction between the two phases is saturated. Moreover, at a constant roasting temperature, excessive ZnF₂ will make the clinker shrink and harden; the chemical activity will be reduced, and the reaction will be insufficient. Therefore, the fluorination rate of rare earths is optimal when the addition of ZnF₂ equals 100%.

3.2.2. Effect of Temperature on the Fluorination Rate of Rare Earths

To investigate the effect of the fluorination rate of rare earths at different roasting temperatures, the conditions of the roasting process were set as follows: the addition of ZnF₂ was 100%, and the roasting time of the material was 90 min. The effect of different roasting temperatures, from 750 to 950 °C, on the fluorination rate of rare earths was investigated, and the results are shown in Figure 4a. As shown in Figure 4a, the rare earth fluorination rate increased significantly from 85.45% to 95.69% as the roasting temperature increased from 750 °C to 850 °C. As the roasting temperature continued to increase, the rare earth fluorination rate decreased. As shown in Figure 4b, the diffraction peaks of NdF₃, CeF₃, PrF₃, and GdF₃ appeared at the roasting temperature of 800 °C, which means that the oxidized rare earths reacted with ZnF₂ at this temperature; however, the fluorination rate of rare earths was lower at this time. The diffraction peaks of the fluorinated rare earths were enhanced at a roasting temperature of 850 °C, while the diffraction peaks of the fluorinated rare earths decreased in intensity at 900 °C. From a kinetic point of view, the increase in roasting temperature accelerates the intermolecular collision, thus accelerating the fluorination reaction. Therefore, the rare earth element fluorination rate increases when the roasting temperature increases from 750 °C to 850 °C. However, as the roasting temperature increases, the clinker begins to shrink and harden, and the collision between molecules is weakened, which is unfavorable to the fluoridation reaction, so, when the roasting temperature reaches 900 °C, the fluorination rate of rare earths decreases. Therefore, the optimal roasting temperature was determined to be 850 °C.

3.2.3. Effect of Roasting Time on Fluorination Rate

To investigate the effect of the rare earth element leaching rate at different roasting times, the conditions of the roasting process of the raw material were set as follows: 100% ZnF₂ addition and 850 °C roasting temperature. The effect of the rare earth elements’ leaching rate at a roasting time of 30–150 min is shown in Figure 5a. From the curve of different roasting times in Figure 5a, it can be seen that the rare earth element fluorination rate increased from 81.21% to 95.69% at a roasting time of 30–90 min, and it tended to be unchanged when the roasting time continued to increase. From the XRD spectra of the clinker with different roasting times in Figure 5b, it can be seen that the peak value of fluorinated rare earths was significantly enhanced when the roasting time increased from 30 to 90 min, and the peak value of fluorinated rare earths essentially remained unchanged when the roasting time continued to be increased. Because the reaction between the oxidized rare earths and ZnF₂ is a solid-phase reaction, prolonging the roasting time can ensure full contact and reaction between the reactants, so, when the roasting time is increased from 30 min to 90 min, the fluorination rate of rare earths increases. The optimal reaction limit between the reactants is reached by increasing the roasting time, so the rare earth elements’ fluorination rate remains unchanged. Considering the cost and economic benefits, 90 min was chosen as the optimal roasting time.

3.3. Box–Behnken Experimental Design

Although RSM cannot explain the response mechanism, the method is able to optimize the test parameters with a minimum number of trials. RSM finds the optimum interval by building a mathematical model and then finds the optimum value in this interval [28]. In RSM, the Box–Behnken design (BBD) method is used to investigate the effect of the interaction of the test factors on the response value [29].

3.3.1. Statistical Analysis and ANOVA

Analysis of variance (ANOVA) is a statistical method used to assess the significance and applicability of regression models [30]. Experiments were conducted based on the BBD principle, and the results are shown in Table 5. Multivariate statistical methods were used to analyze the variance of the fluoridation rate of rare earths, and the results are shown in Table 6. F = 74.60; the signal-to-noise error was only 0.01%, indicating that the constructed regression model is reasonable.

First, the 17 trials were counted using multiple linear regression. Then, the quadratic polynomial (5) was obtained using the least squares method. The normal probability plot shown in Figure 6 shows that the predicted values are in very good agreement with the experimental values, indicating that the model is valid. By calculating the correlation between the actual fluoridation rate and the predicted fluoridation rate (R² = 0.9964), it was verified that the model had a good prediction effect. In the process of building the new mathematical model, factors with little influence as determined by the R² value were removed [31]. The corrected regression equation was fitted to obtain a correlation coefficient of

R_{a d j}^{2}

= 0.9918. In addition, the C.V. was 0.6503%, indicating that the results were reliable. In conclusion, the regression model of the rare earth fluoridation rate established in this paper has a good fitting effect and provides a basis for optimizing the roasting process. The results showed that roasting temperature (A), roasting time (B), and ZnF₂ dosage (C) were the key factors affecting the fluorination rate of rare earths with a confidence level of 95%. Based on this, the interactions between roasting temperature and roasting time, and roasting temperature and ZnF₂ dosage were found to be significant. In addition, A² and B² were found to be significant secondary terms.

Rare earth fluorination rate (%) = 95.98 − 0.763A − 0.8435B + 6.43C − 3.62AB + 2.05AC + 0.0369BC − 3.22A² − 1.61B² − 6.66C²

(5)

3.3.2. Response Surface Optimization Analysis

The interaction between roasting temperature and roasting time is shown in Figure 7a. When the roasting time was short, the fluorination rate of rare earths increased with an increase in the roasting temperature and then tended to stabilize; however, when the roasting time was long, the fluorination rate of rare earths gradually decreased with the increase in roasting temperature. Similarly, when the roasting temperature was low, the fluorination rate of rare earth elements first increased with the increase in roasting time and then stabilized. However, when the roasting temperature was high, the fluorination rate of rare earths gradually decreased with the increase in roasting temperature. Comparing the effects of the two factors, it can be seen that the effects of roasting temperature and roasting time on the fluorination rate of rare earths are the same. In addition, the trend in the rare earth fluorination rate showed a closer relationship between roasting temperature and roasting time. At the appropriate roasting temperature and roasting time, the reaction system will obtain enough thermal energy to accelerate the intermolecular collision, thus promoting the reaction. In contrast, under the condition of higher roasting temperature or longer roasting time, the clinker will contract and harden, thus hindering the diffusion movement between molecules and lowering the rate of fluorination of rare earths. Figure 7b shows a three-dimensional plot of the effect of roasting temperature and ZnF₂ addition on the rare earth fluorination rate. In the range of 750~950 °C, the fluorination rate of rare earths always increased and then decreased slightly with the increase in ZnF₂ addition, which shows that the ZnF₂ addition greatly influenced the fluorination rate of rare earths. When the amount of ZnF₂ added is certain, the fluorination rate of rare earths increases and then decreases with the increase in roasting temperature. The fluorination rate of rare earths is relatively high when the roasting temperature is in the middle of the experimental range. As shown in Figure 7c, the interaction between ZnF₂ dosage and roasting time is not obvious. The effect of increasing the roasting time on the fluorination rate of rare earths was basically unchanged when the amount of ZnF₂ added was fixed. However, when the roasting time was fixed, the fluorination rate of rare earths increased rapidly and then decreased slightly with the increase in ZnF₂ dosage. Therefore, the effect of the roasting time on the rare earth fluorination rate was small compared to that of the ZnF₂ dosage.

Figure 7 shows that, among the three factors of roasting temperature, roasting time, and ZnF₂ addition, the ZnF₂ addition had the most significant effect on the fluorination rate of rare earths, and the fluorination rate of rare earths showed a tendency to rise first and then decrease slightly with the increase in ZnF₂ addition. The experimental results were the same as the law of the one-factor test. This is because, under high-temperature conditions, the rare earths in NdFeB waste can react with ZnF₂ to obtain insoluble fluorinated rare earths. However, when an excessive amount of ZnF₂ is added, sintering occurs, leading to the densification of the clinker structure, slowing down the intermolecular collision effect, which is not conducive to the reaction between ZnF₂ and the rare earth elements, thus reducing the rate of rare earth fluorination.

Using the response surface model, an optimized roasting process was obtained to roast the NdFeB waste with 119% ZnF₂ addition for 91 min at 828 °C, and the rare earth fluorination rate reached 97.29%. The optimized roasting conditions were tested repeatedly with an error of less than 3%.

3.4. Compositional Analysis of Acid Leach Products

The content of mixed rare earth fluorides after hydrochloric acid leaching of the clinker obtained according to the above optimal roasting conditions is shown in Table 7. The content of mixed rare earth fluorides was as high as 98.92%, mainly comprising NdF₃, CeF₄, PrF₃, and GdF₃, and the content of NdF₃ was 50.26%; the content of CeF₄ was 30.97%; the content of PrF₃ was 10.28%; and the content of GdF₃ was 3.04%. The content of these four fluorinated rare earths accounted for 94.55% of the total amount of mixed fluorinated rare earths, which is consistent with the proportion of various rare earth oxides in the raw material, indicating that hydrochloric acid has removed impurities such as iron oxide and zinc oxide from the mixed fluorinated rare earths.

The morphology and structure of the acid-leaching products under the optimal roasting conditions were characterized using XRD and SEM, as shown in Figure 8 and Figure 9. During the Rietveld process, the positions of the atoms were fixed to be the same as that of NdF₃. The results show that the mixed fluorinated rare earth products existed as doped cells rather than as polymorphs, with the chemical formula of Nd_0.53Ce_0.33Pr_0.11Gd_0.03F₃ and the crystal structure of a hexagonal shape. The scanning electron microscope showed that the mixed fluorinated rare earths were irregular and had many holes distributed on the surface, which existed independently. This may be due to the chemical reaction of hydrochloric acid on the material during the acid leaching process, which destroys the impurities such as iron oxide and zinc oxide on the surface and strips them off, leading to the appearance of holes on the surface of the microstructure. Since the main component of the product is fluorinated rare earths and the temperature during the roasting reaction is high, a smooth surface and dense block structure are formed.

3.5. Recommendations for the Process

First, crushed NdFeB scrap is mixed with 119% ZnF₂ and roasted at 828 °C for 91 min. The roasted clinker is washed with ammonia to remove ZnF₂, while filtrate 1 is replenished with ammonia and returned to the washing process, thus recycling the ammonia. The washed slag is leached with 9 M hydrochloric acid to obtain fluorinated rare earth slag, while filtrate 2 is supplemented with hydrochloric acid and returned to the acid leaching process, thus realizing hydrochloric acid recycling. Afterward, the fluorinated rare earth slag is dried, oxidized, and roasted to obtain high-purity mixed rare earth oxide products or subjected to molten salt electrolysis to obtain high-purity mixed rare earth metal products. Therefore, the process shown in Figure 10 realizes the synergistic extraction and comprehensive utilization of NdFeB waste, which has good economic benefits and industrial application value.

4. Conclusions

This study thoroughly studied the reaction mechanism and process conditions of ZnF₂ fluoride recovery from NdFeB waste. The research results of this project will lay the foundation for promoting the application of pyrometallurgical technology in NdFeB waste. The main findings and results obtained are as follows:

The thermodynamic analysis results show that all the rare earth elements in NdFeB waste can react with ZnF₂ to produce fluoride. However, increasing the roasting temperature is unfavorable to the fluoride reaction.
The optimal roasting process was obtained through one-factor experiments: NdFeB waste was added with 100% ZnF₂ and then reacted at 850 °C for 90 min, and the recovery rate of rare earths reached 95.69%. In addition, by analyzing the phase composition of the calcined products, the conclusion of the thermodynamic calculations was verified: an excessive calcination temperature is not conducive to the fluorination reaction, which reduces the recovery of rare earths.
The BBD model was constructed according to the RSM criterion. It was found that the ZnF₂ addition had the greatest effect on the rare earth fluorination rate, followed by roasting temperature and roasting time. In addition, the optimal roasting conditions were determined as follows: a roasting temperature of 828 °C, a roasting time of 91 min, and a ZnF₂ dosage of 119%. The verification experiment demonstrated that the rare earth recovery rate could reach 97.29% under optimal process conditions.
Some aspects remain that require further in-depth research and improvement, including the following: (1) detailed experiments on hydrochloric acid purification of fluorinated rare earths and precise control of leaching conditions, which can increase the efficiency of the subsequent fluorinated rare earth purification work; (2) further research on the separation of fluorinated rare earths, which is conducive to improving the value of the product; and (3) optimization of the process conditions and flow to improve the process reliability and practical applicability.

Author Contributions

Conceptualization, Y.L.; Methodology, Y.L.; Software, Y.L.; Validation, Y.L., Y.Z. and X.L.; Formal analysis, Y.L.; Investigation, Y.L., X.L. and J.W.; Resources, J.W.; Data curation, Y.L.; Writing—original draft, Y.L.; Writing—review & editing, Y.L. and X.L.; Visualization, Y.L.; Supervision, Y.Z., X.L. and J.W.; Project administration, Y.L.; Funding acquisition, X.L. and J.W. All authors have read and agreed to the published version of the manuscript.

Funding

This study was supported by the China Baowu Low Carbon Metallurgy Innovation Foundation (BWLCF202121), Jiangxi Provincial Key Laboratory of Flash Green Development and Recycling (20193BCD40019), academic and technical leaders of major disciplines in Jiangxi Province (20213BCJ22003), the Jiangxi Province Science and Technology Innovation High end Talent Project (jxsq2023201012), the Jiangxi Provincial Natural Science Foundation Youth Fund (20242BAB20194), and the Yichun Science and Technology Plan Project (2023YBKJGG04).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The original contributions presented in the study are included in the article, further inquiries can be directed to the corresponding authors.

Conflicts of Interest

Author Jinliang Wang was employed by the company Jiangxi Chunhua Lithium Industry Co., Ltd. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

References

Orefice, M.; Bulck DV, A.; Blanpain, B.; Binnemans, K. Selective Roasting of Nd-Fe-B Permanent Magnets as a Pretreatment Step for Intensified Leaching with an Ionic Liquid. J. Sustain. Metall. 2020, 6, 91–102. [Google Scholar] [CrossRef]
Du, C.; Ma, S.; Xie, M.; Yang, F.; Zhao, Z.; Chen, Y.; Ma, Y. Recovery of high-value rare earth elements from waste NdFeB by the water-soluble ammonium salt [Hbet]cl. Sep. Purif. Technol. 2023, 308, 122946. [Google Scholar] [CrossRef]
Gebregiorgis, T.A.; Mentore, V.; Duarte, F.C.; Prasad, S.; Rtimi, S. Emerging technologies for the recovery of rare earth elements (REEs) from the end-of-life electronic wastes: A review on progress, challenges, and perspectives. Environ. Sci. Pollut. Res. Int. 2020, 27, 36052–36074. [Google Scholar]
Binnemans, K.; Jones, P.T.; Muller, T.; Yurramendi, L. Rare Earths and the Balance Problem: How to Deal with Changing Markets? J. Sustain. Metall. 2018, 4, 126–146. [Google Scholar] [CrossRef]
Yadav, J.; Sarker, K.S.; Bruckard, W.; Jegatheesan, V.; Haque, N.; Singh, N.; Pramanik, B.K. Greening the supply chain: Sustainable approaches for rare earth element recovery from neodymium iron boron magnet waste. J. Environ. Chem. Eng. 2024, 12, 113169. [Google Scholar] [CrossRef]
Yang, Y.; Lan, C.; Guo, L.; An, Z.; Zhao, Z.; Li, B. Recovery of rare-earth element from rare-earth permanent magnet waste by electro-refining in molten fluorides. Sep. Purif. Technol. 2020, 233, 116030. [Google Scholar] [CrossRef]
Zhang, Y.; Gu, F.; Su, Z.; Liu, S.; Anderson, C.; Jiang, T. Hydrometallurgical Recovery of Rare Earth Elements from NdFeB Permanent Magnet Scrap: A Review. Metals 2020, 10, 841. [Google Scholar] [CrossRef]
Swain, N.; Mishra, S. A review on the recovery and separation of rare earths and transition metals from secondary resources. J. Clean. Prod. 2019, 220, 884–898. [Google Scholar] [CrossRef]
Ji, B.; Li, Q.; Zhang, W. Leaching recovery of rare earth elements from the calcination product of a coal coarse refuse using organic acids. J. Rare Earths 2022, 40, 318–327. [Google Scholar] [CrossRef]
Bian, Y.; Guo, S.; Xu, Y.; Tang, K.; Lu, X.; Ding, W. Recovery of rare earth elements from permanent magnet scraps by pyrometallurgical process. Rare Met. 2022, 41, 1697–1702. [Google Scholar] [CrossRef]
Yin, T.; Xue, Y.; Yan, Y.; Ma, Z.; Ma, F.; Zhang, M.; Wang, G.; Qiu, M. Recovery and separation of rare earth elements by molten salt electrolysis. Int. J. Miner. Metall. Mater. 2021, 28, 899–914. [Google Scholar] [CrossRef]
Liu, B.; Zhu, N.; Li, Y.; Wu, P.; Dang, Z.; Ke, Y. Efficient recovery of rare earth elements from discarded NdFeB magnets. Process Saf. Environ. Prot. 2019, 124, 317–325. [Google Scholar] [CrossRef]
Shirayama, S.; Okabe, T.H. Selective Extraction and Recovery of Nd and Dy from Nd-Fe-B Magnet Scrap by Utilizing Molten MgCl₂. Metall. Mater. Trans. B 2018, 49, 1067. [Google Scholar] [CrossRef]
Venkatesan, P.; Hoogerstraete, T.V.; Binnemans, K.; Sun, Z.; Sietsma, J.; Yang, Y. Selective Extraction of Rare-Earth Elements from NdFeB Magnets by a Room-Temperature Electrolysis Pretreatment Step. ACS Sustain. Chem. Eng. 2018, 6, 9375–9382. [Google Scholar] [CrossRef]
Kumari, A.; Sahu, S.K. A comprehensive review on recycling of critical raw materials from spent neodymium iron boron (NdFeB) magnet. Sep. Purif. Technol. 2023, 317, 123527. [Google Scholar] [CrossRef]
Hua, Z.; Wang, J.; Wang, L.; Zhao, Z.; Li, X.; Xiao, Y.; Yang, Y. Selective Extraction of Rare Earth Elements from NdFeB Scrap by Molten Chlorides. ACS Sustain. Chem. Eng. 2014, 2, 2536–2543. [Google Scholar] [CrossRef]
Mochizuki, Y.; Tsubouchi, N.; Sugawara, K. Selective Recovery of Rare Earth Elements from Dy containing NdFeB Magnets by Chlorination. ACS Sustain. Chem. Eng. 2013, 1, 655–662. [Google Scholar] [CrossRef]
Lorenz, T.; Bertau, M. Recycling of rare earth elements from FeNdB-Magnets via solid-state chlorination. J. Clean. Prod. 2019, 215, 131–143. [Google Scholar] [CrossRef]
Lee, S.E.A.; Sinclai, C.W. The effect of precipitation on recrystallization in a Mg-Nd alloy. Materialia 2020, 10, 100643. [Google Scholar]
Sun, M.; Hu, X.; Peng, L.; Fu, P.; Ding, W.; Peng, Y. On the production of Mg-Nd master alloy from NdFeB magnet scraps. J. Mater. Process. Technol. 2015, 218, 57–61. [Google Scholar] [CrossRef]
Heo, S.G.; Yang, J.Y.; Oh, S.J.; Seo, S.-J.; Lee, M.H.; Park, K.-T. Extraction of rare earth elements from neodymium (NdFeB) magnet scrap using magnesium halides. J. Rare Earths 2024, in press. [Google Scholar] [CrossRef]
Hua, Z.S.; Wang, L.; Wang, J.; Xiao, Y.P.; Yang, Y.X.; Zhao, Z.; Liu, M.J. Extraction of rare earth elements from NdFeB scrap by AlF3–NaF melts. Mater. Sci. Technol. 2015, 31, 1007–1010. [Google Scholar] [CrossRef]
Önal, M.A.R.; Riaño, S. Binnemans, Alkali baking and solvometallurgical leaching of NdFeB magnets. Hydrometallurgy 2020, 191, 105213. [Google Scholar] [CrossRef]
Yang, Y.; Qin, Y.; Wang, T.; Zhang, Y.; Shi, T.; Li, J.; Xu, H.; Xia, Z.; Sun, H.; Zhao, Z. Separation of neodymium from FeNd alloy and the preparation of porous iron alloy by chemical dealloying in molten chlorides: Application to the recovery of NdFeB wastes. Sep. Purif. Technol. 2023, 311, 123185. [Google Scholar] [CrossRef]
Zhou, H.; Liu, Y.; Ma, B.; Wang, C.; Chen, Y. Strengthening extraction of lithium and rubidium from activated α-spodumene concentrate via sodium carbonate roasting. J. Ind. Eng. Chem. 2023, 123, 248–259. [Google Scholar] [CrossRef]
Wang, Y.; Ma, Y.; Lu, Y.; Gao, Q.; Wu, Y.; Wang, Y.; Zhang, C. Phase diagram thermodynamic calculation of KNO₃-NaNO₂-KNO₂ ternary system molten salt and its thermophysical properties investigation for thermal energy storage. J. Energy Storage 2024, 96, 112422. [Google Scholar] [CrossRef]
Cheng, W.; Liu, X.; Tan, C.; Yao, Q.R.; Wang, J.; Rao, G.H.; Zhou, H.Y. Experimental determination and thermodynamic calculation of phase equilibria in the Sm-Fe-B ternary system. Calphad 2024, 85, 102706. [Google Scholar] [CrossRef]
Lee, H.W.; Kim, O.J. Optimization of turbidity removal and floc formation for ballasted flocculation using magnetite-based agents by response surface methodology. Chemosphere 2024, 363, 142932. [Google Scholar] [CrossRef]
Anaklı, D.; Erşan, M. Optimization of reduced graphene oxide yield using response surface methodology. Diam. Relat. Mater. 2024, 148, 111524. [Google Scholar] [CrossRef]
Battestini-Vives, M.; Xiao, X.; Lipnizki, F.; Rudolph-Schöpping, G. Response surface methodology to optimize membrane cleaning in nanofiltration of kraft black liquor. Sep. Purif. Technol. 2025, 354, 128626. [Google Scholar] [CrossRef]
Abeng, D.; Sutaryo, S.; Purnomoadi, A.; Susanto, S.; Purbowati, E.; Adiwinarti, R.; Purwasih, R.; Widiharih, T. Optimization of methane production from dairy cow manure and germinated papaya seeds using response surface methodology. Case Stud. Chem. Environ. Eng. 2024, 10, 100927. [Google Scholar] [CrossRef]

Figure 1. XRD pattern of NdFeB waste.

Figure 2. Plot of the reaction

{Δ G}_{T}^{θ}

versus T for the roasting process.

Figure 2. Plot of the reaction

{Δ G}_{T}^{θ}

versus T for the roasting process.

Figure 3. (a) Effect of the ZnF₂-to-raw material mass ratio on rare earth element fluorination rate; (b) XRD spectra of the clinker with different ZnF₂ dosages.

Figure 4. (a) Effect of roasting temperature on rare earth element fluorination rate; (b) XRD spectra of the clinker with different roasting temperatures.

Figure 5. (a) Effect of roasting time on leaching; (b) XRD spectra of the clinker with different roasting times.

Figure 6. Comparison between predicted and actual values of rare earth element fluorination rates.

Figure 7. Surface plot of the response of the interaction between different factors on the rare earth element fluorination rate: (a) roasting temperature and roasting time, (b) roasting temperature and mass ratio of ZnF₂ to raw material, and (c) mass ratio of ZnF₂ to raw material and roasting time.

Figure 8. XRD graph of leached products.

Figure 9. Scanning electron microscope images of the leachate product at different scales: (a) 1 µm; (b) 500 nm; (c) 200 nm; (d) 100 nm.

Figure 10. Process flow for recovering rare earths from NdFeB waste through the fluorination reaction.

Table 1. Proportions of the rare earth oxides in NdFeB waste.

Component	CeO₂	Pr₆O₁₁	Nd₂O₃	Sm₂O₃	Gd₂O₃	Dy₂O₃	Ho₂O₃	La₂O₃	Eu₂O₃	Tb₂O₃	Er₂O₃	Y₂O₃
Content (wt%)	28.82	10.90	51.57	1.91	3.53	2.29	0.79	0.18	<0.2	<0.2	<0.2	<0.2

Table 2. Other metal elements contained in NdFeB waste and their contents.

Component	Fe *	Al	Co	K	Ca	Zn	Cu	Ni	Na	Mg	Pb	B
Content (wt%)	48.22	0.96	0.50	0.41	0.27	0.22	0.18	0.15	0.13	0.12	0.12	0.09

* Indicates a chemical titration method.

Table 3. Experimental levels and factors design using RSM theory.

Levels	Factors
Levels	Roasting Temperature (°C)	Roasting Time (min)	ZnF₂ Dosage (%)
−1	750	30	50
0	850	90	100
1	950	150	150

Table 4. The main chemical reactions in the roasting process.

No.	Reactions
1	3ZnF₂ + Nd₂O₃ = 2NdF₃ + 3ZnO
2	3ZnF₂ + Ce₂O₃ = 2CeF₃ +3ZnO
3	3ZnF₂ + Gd₂O₃ = 2GdF₃ + 3ZnO
4	3ZnF₂ + La₂O₃ = 2LaF₃ + 3ZnO
5	3ZnF₂ + Ho₂O₃ = 2HoF₃ + 3ZnO
6	3ZnF₂ + Dy₂O₃ = 2DyF₃ + 3ZnO
7	9ZnF₂ + Pr₆O₁₁ = 6PrF₃ + 9ZnO + O₂ (g)
8	3ZnF₂ + Fe₂O₃ = 2FeF₃ + 3ZnO

Table 5. Box–Behnken design matrix and rare earth fluorination rates.

Run	Roasting Temperature (°C)	Roasting Time (min)	ZnF₂ Dosage (%)	Rare Earth Element Fluorination Rate (%)
1	950	150	100	86.52
2	850	150	50	79.98
3	950	30	100	94.79
4	850	90	100	95.96
5	750	90	50	82.82
6	850	90	100	95.64
7	850	90	100	96.14
8	850	90	100	95.47
9	750	30	100	88.53
10	750	90	150	91.43
11	950	90	150	93.47
12	850	30	150	95.36
13	850	150	150	93.08
14	850	90	100	96.69
15	750	150	100	94.75
16	850	30	50	82.41
17	950	90	50	76.65

Table 6. ANOVA table of the response surface quadratic model.

	Sum of		Mean	F	p-Value
Source	Squares	df	Square	Value	Prob > F	Significant
Model	671.37	9	74.60	215.03	<0.0001
A—Roasting temperature (°C)	4.66	1	4.66	13.42	0.0080
B—Roasting time (min)	5.69	1	5.69	16.41	0.0049
C—Mass ratio of ZnF₂ to raw material (%)	331.06	1	331.06	954.29	<0.0001
AB	52.51	1	52.51	151.37	<0.0001
AC	16.83	1	16.83	48.51	0.0002
BC	0.0054	1	0.0054	0.0157	0.9039
A²	43.73	1	43.73	126.04	<0.0001
B²	10.92	1	10.92	31.47	0.0008
C²	186.96	1	186.96	538.92	<0.0001
Residual	2.43	7	0.3469
Lack of fit	1.53	3	0.5091	2.26	0.2236	Not significant
Pure error	0.9011	4	0.2253
Cor total	673.80	16
Fit statistics
Std. Dev.	0.5890		R²		0.9964
Mean	90.57		Adjusted R²		0.9918
C.V.	0.6503		Predicted R²		0.9616
PRESS	25.85		Adeq Precision		42.3543

C.V., coefficient of variation.

Table 7. Mixed rare earth fluoride content.

REF	CeF₄	PrF₃	NdF₃	SmF₃	GdF₃	DyF₃	HoF₃	LaF₃	EuF₃	TbF₃	ErF₃	YF₃	Total
Content (wt%)	30.97	10.28	50.26	1.74	3.04	1.70	0.54	0.18	0.08	0.10	0.01	0.01	98.92

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

Share and Cite

MDPI and ACS Style

Liu, Y.; Zhong, Y.; Lei, X.; Wang, J. Mechanism and Experimental Study on the Recovery of Rare Earth Elements from Neodymium Iron Boron Waste Using the ZnF₂ Fluorination Method. Materials 2024, 17, 5807. https://doi.org/10.3390/ma17235807

AMA Style

Liu Y, Zhong Y, Lei X, Wang J. Mechanism and Experimental Study on the Recovery of Rare Earth Elements from Neodymium Iron Boron Waste Using the ZnF₂ Fluorination Method. Materials. 2024; 17(23):5807. https://doi.org/10.3390/ma17235807

Chicago/Turabian Style

Liu, Youwei, Yuan Zhong, Xiang Lei, and Jinliang Wang. 2024. "Mechanism and Experimental Study on the Recovery of Rare Earth Elements from Neodymium Iron Boron Waste Using the ZnF₂ Fluorination Method" Materials 17, no. 23: 5807. https://doi.org/10.3390/ma17235807

APA Style

Liu, Y., Zhong, Y., Lei, X., & Wang, J. (2024). Mechanism and Experimental Study on the Recovery of Rare Earth Elements from Neodymium Iron Boron Waste Using the ZnF₂ Fluorination Method. Materials, 17(23), 5807. https://doi.org/10.3390/ma17235807

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Menu

Mechanism and Experimental Study on the Recovery of Rare Earth Elements from Neodymium Iron Boron Waste Using the ZnF₂ Fluorination Method

Abstract

1. Introduction