Abstract
Natural Zn reserves might be exhausted in the following decades. Electric arc furnace (EAF) dust is an essential secondary Zn resource. The recovery of Zn from EAF dust is a critical factor in the sustainable development of the Zn industry and steelmaking industry. Efficient destruction of the ZnFe2O4 crystal structure and separation of Zn and Fe are the key issues to recycle EAF dust. Molten metal chlorides are believed to be effective in breaking the structure of ZnFe2O4. This work clarifies the distinct reaction behavior between ZnFe2O4 and MgCl2 or CaCl2. Experiments show that although MgCl2 is a more powerful chlorinating agent, the capability of separating Zn and Fe is not as good as that of CaCl2. Fast and selective chlorination of Zn from ZnFe2O4 happens by reacting ZnFe2O4 and MgCl2–CaCl2 mixture. CaCl2 facilitates the formation of stable Fe-bearing phases, such as MgFe2O4 and Ca2Fe2O5, thereby inhibiting the chlorination of Fe. Moreover, it becomes more effective in promoting the selectivity of chlorinating Zn by increasing temperature after adding CaCl2 in MgCl2. When ZnFe2O4 is subjected to chlorination using a MgCl2–CaCl2 mixture (MgCl2–CaCl2–ZnFe2O4 molar ratio = 1.5:1.5:1) at 950 °C for 120 min, the process results in a chlorination percentage of over 90% for Zn, with Fe chlorination percentage lower than 5%. These results demonstrate the prospect of efficient separation of Zn from ZnFe2O4 by reaction with molten MgCl2–CaCl2 mixture.
Graphical Abstract
Similar content being viewed by others
Introduction
Zn is an essential element for modern society. The steel industry is the largest consumer of Zn. In 2019, the quantity of Zn used for galvanizing steel was 10,167 kt, which represented over 50% of the global Zn production [1]. Coating a thin Zn layer on the steel surface is a simple and efficient way to protect the base metal from corrosion and rusting. Galvanized steel is thus among the most popular steel types and is widely applied in various industrial sectors. When scrap steel is remelted in an electric arc furnace (EAF), all Zn evaporates and gets into the EAF dust. In these cases, the Zn content in EAF dust varies from 5 to 40% [2]. Figure 1 shows the schematic illustration of the mass flow of Zn in the steel industry. Complete recycling of Zn from EAF dust would increase the circularity of Zn and reduce the dependence on natural resources. However, the actual situation is that a large portion of EAF dust (around 37%) is sent to landfills because of the technical difficulties in recycling. Rostek et al. estimated that the loss of Zn in the steel industry was 500 kt in 2019 [1]. On the other hand, Zn is recognized as one of the elements with limited natural supplies. The US Geological Survey’s latest report suggested that the Zn mine might be exhausted within 20 years [3]. Due to the value of Zn and the gradual depletion of natural reserves, boosting the recycling ratio of Zn from EAF dust has become a vital issue in Zn sustainability.
ZnFe2O4, ZnO, and Fe2O3 are the main components in EAF dust. The difficulty in recovering Zn from EAF dust is because of the high chemical stability of ZnFe2O4 [4, 5]. Both pyrometallurgical and hydrometallurgical techniques can treat EAF dust [6,7,8,9,10,11,12,13,14]. Pyrometallurgy is the primary choice due to the high metal recovery ratio and relatively short process [15]. The Waelz process and the rotary hearth furnace technology are two typical routes with commercial applications. Both are based on carbothermic reduction at high temperatures [16]. The Zn-bearing oxides in the dust are reduced by coke at around 1200 °C to form Zn vapor, which is later re-oxidized by air to ZnO. The final recycled product is crude ZnO powder instead of metallic Zn. However, high-energy consumption and massive emission of CO2 are the main drawbacks, especially as the goal of achieving carbon neutrality has become an international consensus. Therefore, innovations in Zn recycling from EAF dust are urgently needed. Recently, Nagasaka et al. proposed a “CaO addition process” to replace the Waelz process after years of fundamental and systematic research input [17]. In this technology, EAF dust is mixed with CaO at a high temperature (e.g., 1000 °C). CaO would react with ZnFe2O4 in the dust to form ZnO and Ca2Fe2O5. Because ZnO is far easier to dissolve than ZnFe2O4 in acid or alkali solutions, metallic Zn can be obtained after leaching and electrolysis [18, 19]. The most significant advantage of this technology is that carbothermic reduction is no longer necessary. However, the slow solid-state reaction between dust and CaO remains a challenge for industrial applications.
Chlorination-based pyrometallurgical processes have significant advantages in terms of efficiency and low carbon emissions [20,21,22,23,24,25]. Zn-bearing oxides can be converted into ZnCl2 by Cl2, polyvinyl chloride (PVC), or metal chlorides through a rapid gas–solid or liquid–solid reaction. Metallic Zn can be easily obtained by electrolyzing molten ZnCl2. In principle, no carbon-based reducing agent is required for the chlorination process. Matsuura et al. [26,27,28] carried out systematic research on using Cl2 as the chlorinating agent. As reported, Zn in EAF dust could be efficiently and selectively chlorinated utilizing a mixture of Cl2 and O2. However, the high corrosiveness and toxicity of Cl2 pose challenges to equipment corrosion and safety issues. Finding a Cl2-free substance as the chlorinating agent to make the process easier and safer is necessary. Apart from Cl2, PVC has been utilized as a functional chlorinating additive to assess the feasibility of recycling EAF dust for metal extraction [29,30,31,32,33,34]. By reacting EAF dust with PVC, Zn in EAF dust is converted to ZnCl2 by hydrogen chloride (HCl) gas evolved from PVC. However, generating dioxin and other pollutants from PVC at high temperatures is a disadvantage.
Molten metal chlorides can also be used as chlorinating agents in place of Cl2 and PVC. Guo et al. clarified that molten CaCl2 converts Zn in ZnFe2O4 into gaseous ZnCl2 while making Fe remain in the solid residue as CaFe2O4 [35,36,37,38]. This result indicates a high selectivity of chlorinating Zn by CaCl2. However, the sluggish reaction is not satisfied from a practical standpoint. Since MgCl2 is another common chloride with higher reactivity to chlorinate solid oxides [39, 40], we propose an alternative technology of using MgCl2 to treat EAF dust. The target product is ZnCl2. Previously, we demonstrated the capability of MgCl2 to chlorinate ZnFe2O4 [41]. MgCl2 converts Zn in ZnFe2O4 to ZnCl2 more efficiently than CaCl2. Reaction temperature could be lower, and the rate is larger when using MgCl2. However, a part of Fe in ZnFe2O4 is also chlorinated to form volatile iron chlorides, which means MgCl2 is not as good as CaCl2 in terms of selective chlorination of Zn.
Considering the distinct features of MgCl2 and CaCl2, using MgCl2 with the addition of CaCl2 may lead to fast and selective chlorination of Zn from ZnFe2O4. This work compares the chlorination behaviors of ZnFe2O4 between using MgCl2 and using CaCl2 and demonstrates the superior efficacy of using the MgCl2–CaCl2 mixture. The reaction mechanism and the effects of thermodynamic conditions (mixture composition, temperature, etc.) on the chlorination behaviors of Zn and Fe are clarified. This work may optimize the technology based on utilizing chlorides for Zn recovery from EAF dust.
Materials and Method
Materials
Reagent grade ZnO (99%, Aladdin), Fe2O3 (99%, Aladdin), MgCl2 (99%, Aladdin), and CaCl2 (97%, Aladdin) were used. ZnO and Fe2O3 (molar ratio of 1:1) were thoroughly mixed and sintered in air at 1100 °C for 24 h to prepare ZnFe2O4. The as-prepared ZnFe2O4 was ground, and particles in size range from 0.10 mm to 0.28 mm were used in the experiments. All chemicals were vacuum dried at 200 °C for at least 24 h before use.
Experimental Procedure
The experiment was performed in a horizontal tube furnace, as shown in Fig. 2. A MgO crucible (cuboid shape, length 60 mm, width 30 mm, depth 20 mm) containing a powdery sample was set in a quartz tube in the furnace. The samples were a mixture of MgCl2 and ZnFe2O4 (0.2856 g and 0.2410 g, molar ratio = 3:1), a mixture of CaCl2 and ZnFe2O4 (0.3329 g and 0.2410 g, molar ratio = 3:1), and mixtures of MgCl2, CaCl2, and ZnFe2O4 with different molar ratios. High-purity Ar was introduced as the protective gas at 200 mL min−1. The reaction temperature was from 800 °C to 1000 °C.
The weight loss of the sample during the reaction was measured. As shown in Fig. 2, a corundum tube enclosing a thermocouple was placed inside the quartz tube, ensuring that the tube tip was at the center (hot zone). Once the furnace temperature reached the desired value, the reactant-filled MgO crucible was moved from the cold zone to the hot zone within the quartz tube. After a specific duration, the corundum tube was quickly pushed to return the crucible back to the quartz tube end (cold zone). The crucible was then retrieved using a stainless-steel clamp and weighed. To prevent the direct contact between the hot crucible and the balance, a quartz crucible was placed on the balance. After weighing, the crucible was placed back in the quartz tube center. These steps were repeated multiple times to plot the weight loss curve.
During reaction, volatiles would form, evaporate, and deposit on the inner quartz wall in the cold zone, while solid residues should remain in the crucible. In some cases, quartz wools were set in the quartz tube to collect the volatiles. After reaction, the quartz tube was rinsed with distilled water to dissolve the volatiles deposited on the inner wall. The crucible was also rinsed by immersing it in distilled water. Accordingly, substances after reaction were divided into four parts (Fig. 2): Rinse Solution-V (containing soluble phases in volatiles denoted by V) and Solids-V (insoluble phases in volatiles) from volatiles, Rinse Solution-R (containing soluble phases in residues in crucible denoted by R) and Solids-R (insoluble phases in residues in crucible) from the residues in crucible. These four parts of substances were analyzed by phase or chemical analysis, respectively.
Analysis and Characterization
Phase compositions of Solids-V and Solids-R were analyzed by an X-ray diffractometer (XRD, Cu Kα, 40 kV, 40 mA, 5° min−1, Bruker, D8 Advance). When using quartz wools to collect volatiles, the quartz wools were ground into powder in a glove box after the reaction. The powdery sample was also analyzed by XRD. In addition, the microstructures of Solids-V and Solids-R were observed by a scanning electron microscope (SEM, Zeiss Gemini 450) equipped with an energy-dispersive spectrometer (EDS, Oxford instrument). Raman spectra were recorded on a spectrometer (WITec, Alpha300R) using a laser source with an excitation wavelength of 532 nm in the range of 100–900 cm−1. Rinse Solution-V and Rinse Solution-R were subject to chemical analysis by an inductively coupled plasma mass spectrometer (ICP − MS, ThermoFisher, iCAP RQ). The fractions of Zn and Fe (FM) in Rinse Solution-V and Rinse Solution-R can be calculated by
where CM is the concentration of a metal ion in the solution (g L−1), V is the total volume of the solution (L), and WM is the total mass of Zn or Fe in initial ZnFe2O4. Although deposits on the quartz tube wall may contain water-insoluble FeOCl that is difficult to recover completely [41], this phase was formed in tiny quantities. Therefore, the sum of the fractions of Zn or Fe in Rinse Solution-V and Rinse Solution-R is defined as the chlorination percentage.
Results and Discussion
Distinct Reaction Behavior of ZnFe2O4 Between Using MgCl2 and Using CaCl2
Based on our previous experiments, when the molar ratio of MgCl2 to ZnFe2O4 was 1:1, only less than 50% of Zn in ZnFe2O4 was chlorinated [41]. Therefore, a molar ratio of 3:1 was adopted to ensure sufficient chlorination of Zn in this work. Figure 3 shows the weight loss percentage curves of pure chlorides and chloride-oxide mixtures at 950 °C in Ar. The weight loss rate of chloride/oxide mixtures is more significant than their corresponding pure chlorides, suggesting the chlorination of ZnFe2O4 by MgCl2 or CaCl2 to form volatiles. The weight of MgCl2–ZnFe2O4 mixture decreases rapidly and reaches a constant value within 10 min, while the weight loss of CaCl2–ZnFe2O4 mixture is much slower. The result confirms that MgCl2 is more potent in chlorination compared with CaCl2.
Figure 4a shows the variation of chlorination percentage of Zn and Fe for the MgCl2–ZnFe2O4 and CaCl2–ZnFe2O4 mixtures (molar ratio = 3:1 for both cases) at 950 °C in Ar. According to the increasing rate of chlorination percentage, chlorination of Zn is faster than that of Fe in both cases, suggesting the preference for chlorinating Zn over Fe. In the case of using MgCl2, chlorination of Zn shows a tendency of “fast followed by slow,” while chlorination of Fe is “slow followed by fast.” As a result, the molar ratio of Zn to Fe in Rinse Solution-V (the solution after rinsing volatiles with water) is gradually declining (Fig. 4b). In the case of using CaCl2, chlorination of Zn is slower, and it seems like that Fe is rarely chlorinated. Accordingly, a continuous increase in the molar ratio of Zn to Fe in Rinse Solution-V can be noticed. The ratio reaches more than 100 after reaction for 10 min, which is far larger than the values of using MgCl2. These results confirm that although MgCl2 is a potent chlorinating agent, the capability of separating Zn and Fe is not as good as that of CaCl2.
Figure 5 shows the XRD patterns of Solids-R (insoluble phases in residues in the crucible) after reacting ZnFe2O4 with MgCl2 or CaCl2 at 950 °C for 120 min. When reacting with MgCl2, a spinel phase and MgO were detected. The spinel phase may be incompletely reacted ZnFe2O4 and newly formed MgFe2O4. When reacting with CaCl2, CaFe2O4 and Ca2Fe2O5 were detected. ZnFe2O4 should also remain in the CaCl2 system. However, the presence of products such as CaFe2O4 and Ca2Fe2O5 may envelop the remaining ZnFe2O4, making it challenging to detect through XRD. Figure 6 shows the SEM images and EDS mapping results of Solids-R in both cases. Fine particles can be observed in the products of the reaction between ZnFe2O4 and MgCl2. EDS mapping indicates high concentrations of Mg and O while low concentrations of other elements in these particles (Fig. 6a). Although a small amount of MgFe2O4 forms, most of the solid phase products are MgO. In the case of reacting with CaCl2, strips several hundred microns long can be detected. They were identified as Ca2Fe2O5 (Fig. 6b). These observations are consistent with the previous findings of Guo et al. [37, 38].
We have clarified the reaction behavior between solid ZnFe2O4 and molten MgCl2 in a previous work [41]. The chlorination of Zn is with higher priority
Although the priority is lower, the chlorination of Fe also occurs
In the case of reacting with CaCl2, a similar reaction will occur
Although reaction (4) is not spontaneous at standard state conditions, deviation from the standard state of products (CaO and ZnCl2) may cause the reaction to proceed forward. Figure 7 shows the pH of the water-rinsed solutions of Solids-R. The increase in the pH after reacting with CaCl2 is due to hydration of newly formed CaO, while the decrease in the pH after reacting with MgCl2 is due to the hydration of MgO.
The dissolution behavior of CaO in molten CaCl2 and that of MgO in molten MgCl2 are rather different, which might be the primary reason for the dissimilar performances in separating Zn and Fe between the two chlorides. Because CaO is highly soluble and dissociable in molten CaCl2 (solubility > 20 mol% at 950 °C [42]), CaO formed in reaction (4) will immediately dissociate into Ca2+ and O2− in molten CaCl2 and react with Fe2O3 to form new compounds by
CaFe2O4 and Ca2Fe2O5 are more stable than Fe2O3 [43], and thus, the chlorination of Fe is inhibited.
In contrast, O.2− is almost insoluble in molten MgCl2 [44]. Although the following reactions may turn Fe2O3 into more stable MgFe2O4
, the reaction rates must be low due to the low concentration of O2− or limited surface area for reaction. As evidenced by the experimental results (Fig. 6), there was only a tiny amount of MgFe2O4 in Solids-R after reacting ZnFe2O4 with MgCl2, and the particle size is far smaller than that of Ca2Fe2O5 in the case of using CaCl2. As a result, a considerable part of Fe was chlorinated to FeCl3 when using MgCl2.
Therefore, the mechanisms of chlorinating ZnFe2O4 by MgCl2 and CaCl2 can be summarized, as illustrated in Fig. 8. When MgCl2 is used as the chlorinating agent, Zn in the lattice of ZnFe2O4 is first chlorinated to form ZnCl2, MgO, and Fe2O3 [reaction (2)]. The newly formed MgO and Fe2O3 in contact with each other would react to produce a tiny amount of MgFe2O4. As Zn gets consumed, MgCl2 gradually reacts with Fe2O3 to form FeCl3 [reaction (3)]. Both ZnCl2 and FeCl3 leave the melt as volatiles soon after their formation. The final products remaining in the system as solid residues are MgO and MgFe2O4. In the case of using CaCl2, after the reaction between CaCl2 and ZnFe2O4 [reaction (4)], the newly formed CaO immediately dissolves in the melt and dissociates into mobile O2−. A concentration gradient layer of O2− from the interface to the bulk will form. The melt film surrounding the newly formed Fe2O3 will instantly become rich in O2−, which turns the surface of Fe2O3 into more stable compounds such as Ca2Fe2O5 and inhibits the further chlorination of Fe [reactions (5)–(7)]. Ca2Fe2O5 is the major solid phase remaining in the system after sufficient reaction.
These findings suggest that the superior performance of CaCl2 in separating Zn from Fe is because of the highly soluble and mobile nature of O2− in molten CaCl2. Therefore, we consider that adding CaCl2 in MgCl2 may promote the mobility of O2− in the melt and thus achieve fast and selective separation of Zn from ZnFe2O4.
Chlorination of ZnFe2O4 by MgCl2–CaCl2
Figure 9 shows the chlorination percentages of Zn and Fe and molar ratios of Zn to Fe in Rinse Solution-V after the reactions between ZnFe2O4 and different MgCl2–CaCl2 mixtures at 950 °C for 120 min in Ar. The chlorination percentages of Fe decline clearly with the increase of CaCl2 addition. The result demonstrates that adding CaCl2 can enhance the selectivity of chlorinating Zn. Figure 10 shows the variation of chlorination percentages of Zn and Fe during the reactions between ZnFe2O4 and different MgCl2–CaCl2 mixtures at 950 °C in Ar. A decrease in chlorination rate with the increase in CaCl2 addition can be noticed. It is considered that MgCl2 plays the role of fast chlorination. CaCl2 can inhibit the chlorination of Fe, yet it may slow down the overall reaction. There must be an optimum composition of the MgCl2–CaCl2 mixture to achieve the best chlorinating performance by balancing these two effects. It appears that the mixture of MgCl2–CaCl2 = 1:1 performs better than the others.
Volatiles formed during the reaction between ZnFe2O4 and MgCl2–CaCl2 mixture or MgCl2 were collected and characterized by XRD, as shown in Fig. 11. Due to the addition of CaCl2, the diffraction peaks of FeCl3 are obviously weakened, consistent with the chemical analysis results that the addition of CaCl2 lowers the chlorination percentage of Fe (Fig. 10).
As shown in Fig. 12, both the mass and mass fraction of MgCl2 and CaCl2 in the volatiles are analyzed. The results indicate that the volatilization loss of MgCl2 is less than 2.5%, and that of CaCl2 is below 1% in all cases. Due to the limited amount, CaCl2 in the volatiles is difficult to detect by XRD. Additionally, the addition of CaCl2 can further reduce the volatilization loss of MgCl2. Therefore, the volatilization of either MgCl2 or CaCl2 would not cause significant problems for the chlorination of ZnFe2O4.
Figure 13 shows XRD patterns and Raman spectra of Solids-R (insoluble phases in residues in the crucible) after reacting ZnFe2O4 with different MgCl2–CaCl2 mixtures for 120 min at 950 °C in Ar. Diffraction peaks belonging to a spinel phase and MgO are clearly observed in the XRD pattern. Ca2Fe2O5 can also be detected, although the diffraction peaks are weak. The obtained Raman spectra have been deconvoluted into individual Lorentzian peaks for analysis. The peak of T2g mode at ~ 210 cm−1 and the peak of A1g mode at ~ 700 cm−1 indicate the occupation of Mg2+ in the octahedral and tetrahedral sites in the spinel structure [45, 46]. Thus, we infer that the spinel phase in Solids-R should be MgFe2O4 rather than ZnFe2O4 or Fe2O3. Figure 14 shows the SEM image and EDS mapping results of Solids-R (insoluble phases in residues in the crucible) after reacting ZnFe2O4 with MgCl2–CaCl2 mixtures for 120 min at 950 °C in Ar. Phases with compositions close to that of MgFe2O4 (Point A) and MgO (Point B) can be observed. We also notice that with the increase in CaCl2 addition, more MgFe2O4 is generated. These results suggest that the addition of CaCl2 leads to the formation of MgFe2O4 and Ca2Fe2O5, which inhibit the chlorination of Fe.
It is considered that CaCl2 in MgCl2 influences the reaction with ZnFe2O4 from two aspects. At first, since CaCl2 and MgCl2 are perfectly miscible, the addition of CaCl2 is equivalent to lowering the concentration and thermodynamic activity of MgCl2. Assuming reactions (2) and (3) are one-order reactions, the forward reaction rate (r) can be written as r = k[MgCl2], where k is the rate constant and [MgCl2] is the concentration of MgCl2. As the equation shows, a decrease in the concentration of MgCl2 will slow the reaction. On the other hand, the mobility of O2− in the molten salt may rise with the increase in CaCl2 content. Figure 15 illustrates the mechanism. Solid ZnFe2O4 first reacts with MgCl2 in the MgCl2–CaCl2 mixture to produce ZnCl2, MgO, and Fe2O3. A concentration gradient layer of MgCl2 from the interface to the bulk will form because of the fast consumption of MgCl2. Deficiency of MgCl2 at the interface favors the reaction between ZnFe2O4 and CaCl2 to produce mobile O2−. Since the diffusion of MgCl2 from the bulk to the interface promptly refill the deficiency, O2− and Mg2+ turn Fe2O3 into more stable MgFe2O4 via reaction (8). When most MgCl2 is consumed, a tiny amount of CaFe2O5 is also formed via reaction (7). MgO, MgFe2O4, and Ca2Fe2O5 are the solid phases remaining in the system after sufficient reaction. As a result, chlorination of Fe is inhibited, and the selectivity of Zn chlorination is improved.
Synergistic Effect of Temperature and Adding CaCl2
Figure 16a shows the chlorination percentages of Zn and Fe after the reactions between ZnFe2O4 and MgCl2 or MgCl2–CaCl2 mixture for 45 min in Ar at different temperatures. When using MgCl2 only (MgCl2–ZnFe2O4 molar ratio = 3:1), the chlorination percentage of Zn is mainly independent of temperature, while that of Fe decreases slightly as the temperature increases. In the case of using MgCl2–CaCl2 mixture (MgCl2–CaCl2–ZnFe2O4 molar ratio = 2.5:0.5:1), the chlorination percentage of Zn increases from 81 to 94%, and that of Fe decreases from 38 to 15%, as the temperature increases from 800 to 1000 °C. These results indicate that it becomes more effective to promote the selectivity of chlorinating Zn by increasing temperature after adding CaCl2 in MgCl2. Molar ratios of Zn to Fe in Rinse Solution-V in the case of using MgCl2–CaCl2 mixture at 950 °C and 1000 °C are larger than the values using MgCl2 only (Fig. 16b). Figure 16c shows the variation of chlorination percentages of Zn and Fe during the reactions between ZnFe2O4 and MgCl2–CaCl2 mixtures (MgCl2–CaCl2–ZnFe2O4 molar ratio = 2.5:0.5:1) at different temperatures in Ar. A higher temperature favors the chlorination of Zn in terms of amount and rate, while the dependency of the chlorination rate of Fe on temperature is not apparent. Furthermore, the chlorination percentages of Fe exhibit fluctuations that are particularly pronounced at higher temperatures. Our previous study [41] suggested that these fluctuations may be attributed to the formation of insoluble FeOCl from FeClx in the volatiles at elevated temperatures. This reaction may also facilitate the separation of Zn and Fe. Based on the selectivity and rate of chlorination, the ideal temperature for chlorination is 950 °C when using the MgCl2–CaCl2 mixture.
Figure 17 shows the XRD patterns and Raman spectra of Solids-R (insoluble phases in residues in crucible) after reacting ZnFe2O4 with MgCl2–CaCl2 mixture (MgCl2–CaCl2–ZnFe2O4 molar ratio = 2.5:0.5:1) at different temperatures for 45 min in Ar. In all cases, formation of MgFe2O4 can be confirmed. As can be noticed, the diffraction peaks at 35.4° belonging to MgFe2O4 and the Raman spectra peak of A1g mode at ~ 700 cm−1 are significantly intensified as temperature increases. This trend indicates that a higher temperature promotes the formation of MgFe2O4.
The formation of MgFe2O4 from Fe2O3 can be expressed by reaction (8). The formation rate can be written as r = k[Mg2+][O2−]. The rate constant k would increase with temperature, and the concentration of O2− ([O2−]) will also rise due to the increase in solubility. Therefore, the formation rate of MgFe2O4 increases with the temperature. Fast conversion of Fe2O3 into more stable phases, such as MgFe2O4, in molten MgCl2–CaCl2 at higher temperatures could explain why the chlorination of Fe is precluded.
Conclusions
This work clarifies the distinct reaction behavior of ZnFe2O4 between using MgCl2 and using CaCl2. Although MgCl2 is a more powerful chlorinating agent, the capability of separating Zn and Fe is not as good as that of CaCl2. Fast and selective chlorination of Zn from ZnFe2O4 happens by reacting ZnFe2O4 and MgCl2–CaCl2 mixture. By raising the mobility of O2− in the molten salt, CaCl2 facilitates the formation of stable Fe-bearing phases such as MgFe2O4 and Ca2Fe2O5, thereby inhibiting the chlorination of Fe. Moreover, it becomes more effective in promoting the selectivity of chlorinating Zn by increasing temperature after adding CaCl2 in MgCl2. These results demonstrate the prospect of efficient separation of Zn from ZnFe2O4 by reaction with molten MgCl2–CaCl2 mixture.
Change history
23 August 2023
A Correction to this paper has been published: https://doi.org/10.1007/s40831-023-00734-w
References
Rostek L, Espinoza LT, Goldmann D, Loibl A (2022) A dynamic material flow analysis of the global anthropogenic zinc cycle: providing a quantitative basis for circularity discussions. Resour Conserv Recy 180:106154
Ye Q, Li G, Peng Z, Augustine R, Pérez MD, Liu Y, Liu M, Rao M, Zhang Y, Jiang T (2020) Microwave-assisted self-reduction of EAF dust-biochar composite briquettes for production of direct reduced iron. Powder Technol 362:781–789
U.S. Geological Survey (2022) Mineral commodity summaries. U.S. Geological Survey, Washington
Havlík T, e Souza BV, Bernardes AM, Schneider IAH, Miškufová A (2006) Hydrometallurgical processing of carbon steel EAF dust. J Hazard Mater 135:311–318
Wu C, Chang F, Chen W, Tsai M, Wang Y (2014) Reduction behavior of zinc ferrite in EAF-dust recycling with CO gas as a reducing agent. J Environ Manag 143:208–213
Halli P, Agarwal V, Partinen J, Lundström M (2020) Recovery of Pb and Zn from a citrate leach liquor of a roasted EAF dust using precipitation and solvent extraction. Sep Purif Technol 236:116264
Halli P, Hamuyuni J, Leikola M, Lundström M (2018) Developing a sustainable solution for recycling electric arc furnace dust via organic acid leaching. Miner Eng 124:1–9
Kerry T, Peters A, Georgakopoulos E, Dugulan I, Meijer K, Hage J, Offerman E, Yang Y (2022) Zinc vaporization and self-reduction behavior of industrial waste residues for recycling to the HIsarna furnace. J Sustain Metall 8:658–672
Omran M, Fabritius T, Heikkinen E-P (2019) Selective zinc removal from electric arc furnace (EAF) dust by using microwave heating. J Sustain Metall 5:331–340
Tang H, Peng Z, Wang L, Anzulevich A, Rao M, Li G (2023) Direct conversion of electric arc furnace dust to zinc ferrite by roasting: effect of roasting temperature. J Sustain Metall 9:1–12
Van Winkel S, Scheunis L, Verhaeghe F, Blanpain B, Malfliet A (2019) Chlorine addition to existing zinc fuming processes: a thermodynamic study. J Sustain Metall 5:538–550
Wang L, Peng Z, Lin X, Ye Q, Ye L, Zhang J, Liu Y, Liu M, Rao M, Li G (2021) Microwave-intensified treatment of low-zinc EAF dust: a route toward high-grade metallized product with a focus on multiple elements. Powder Technol 383:509–521
Ye Q, Peng Z, Li G, Lee J, Liu Y, Liu M, Wang L, Rao M, Zhang Y, Jiang T (2019) Microwave-assisted reduction of electric arc furnace dust with biochar: an examination of transition of heating mechanism. ACS Sustain Chem Eng 7:9515–9524
Zhang D, Ling H, Yang T, Liu W, Chen L (2019) Selective leaching of zinc from electric arc furnace dust by a hydrothermal reduction method in a sodium hydroxide system. J Clean Prod 224:536–544
Wang J, Zhang Y, Cui K, Fu T, Gao J, Hussain S, AlGarni TS (2021) Pyrometallurgical recovery of zinc and valuable metals from electric arc furnace dust—a review. J Clean Prod 298:126788
Pickles C (2008) Thermodynamic analysis of the selective carbothermic reduction of electric arc furnace dust. J Hazard Mater 150:265–278
Chairaksa-Fujimoto R, Inoue Y, Umeda N, Itoh S, Nagasaka T (2015) New pyrometallurgical process of EAF dust treatment with CaO addition. Int J Min Met Mater 22:788–797
Chairaksa-Fujimoto R, Maruyama K, Miki T, Nagasaka T (2016) The selective alkaline leaching of zinc oxide from Electric Arc Furnace dust pre-treated with calcium oxide. Hydrometallurgy 159:120–125
Miki T, Chairaksa-Fujimoto R, Maruyama K, Nagasaka T (2016) Hydrometallurgical extraction of zinc from CaO treated EAF dust in ammonium chloride solution. J Hazard Mater 302:90–96
Gaballah I, Djona M, Mugica J, Solozobal R (1994) Valuable metals recovery from spent catalysts by selective chlorination. Resour Conserv Recy 10:87–96
Pickles C (2009) Thermodynamic analysis of the selective chlorination of electric arc furnace dust. J Hazard Mater 166:1030–1042
Hu X, Jiang P, Yan Z, Zhu L, Chou K-C, Matsuura H, Tsukihashi F (2013) Selective chlorination reaction of Cu2O and FeO mixture by CaCl2. ISIJ Int 53:541–543
Ma Y, Zhou X, Tang J, Liu X, Gan H, Yang J (2021) One-step selective recovery and cyclic utilization of valuable metals from spent lithium-ion batteries via low-temperature chlorination pyrolysis. Resour Conserv Recy 175:105840
Huang Y, Shao P, Yang L, Zheng Y, Sun Z, Fang L, Lv W, Yao Z, Wang L, Luo X (2021) Thermochemically driven crystal phase transfer via chlorination roasting toward the selective extraction of lithium from spent LiNi1/3Co1/3Mn1/3O2. Resour Conserv Recy 174:105757
Xing Z, Cheng G, Yang H, Xue X, Jiang P (2020) Mechanism and application of the ore with chlorination treatment: a review. Miner Eng 154:106404
Matsuura H, Hamano T, Tsukihashi F (2006) Chlorination kinetics of ZnFe2O4 with Ar-Cl2-O2 gas. Mater Trans 47:2524–2532
Matsuura H, Hamano T, Tsukihashi F (2006) Removal of Zn and Pb from Fe2O3–ZnFe2O4–ZnO–PbO mixture by selective chlorination and evaporation reactions. ISIJ Int 46:1113–1119
Matsuura H, Tsukihashi F (2006) Chlorination kinetics of ZnO with Ar-Cl2-O2 gas and the effect of oxychloride formation. Metall Mater Trans B 37:413–420
Al-Harahsheh M (2018) Thermodynamic analysis on the thermal treatment of electric arc furnace dust-PVC blends. Arab J Sci Eng 43:5757–5769
Al-Harahsheh M, Al-Nu’airat J, Al-Otoom A, Al-jabali H, Al-zoubi M (2019) Treatments of electric arc furnace dust and halogenated plastic wastes: a review. J Environ Chem Eng 7:102856
Al-Harahsheh M, Al-Otoom A, Al-Jarrah M, Altarawneh M, Kingman S (2018) Thermal analysis on the pyrolysis of tetrabromobisphenol A and electric arc furnace dust mixtures. Metall Mater Trans B 49:45–60
Al-Harahsheh M, Kingman S, Al-Makhadmah L, Hamilton IE (2014) Microwave treatment of electric arc furnace dust with PVC: dielectric characterization and pyrolysis-leaching. J Hazard Mater 274:87–97
Altarawneh S, Al-Harahsheh M, Dodds C, Buttress A, Kingman S (2022) Thermal degradation kinetics of polyvinyl chloride in presence of zinc oxide. Thermochim Acta 707:179105
Lee GS, Song YJ (2007) Recycling EAF dust by heat treatment with PVC. Miner Eng 20:739–746
Iwase G, Okumura K (2021) Nonisothermal investigation of reaction kinetics between electric arc furnace dust and calcium chloride under carbon-containing conditions. ISIJ Int 61:2483–2489
Sato H, Okumura K (2020) Recovery of zinc by reaction between electric arc furnace dust and calcium chloride. Tetsu to Hagane 106:82–90
Guo T, Hu X, Hou X, Matsuura H, Tsukihashi F, Zhou G (2011) Chlorination reaction mechanism between ZnFe2O4 and CaCl2. J Univ Sci Technol Beijing 33:474
Guo T, Hu X, Shu Q, Zhou G (2008) Removal of zinc from ZnFe2O4 by selective chlorination and evaporation. J Chin Rare Earth Soc 26:849–853
Cooper EI, Kohn DH (1983) The use of molten magnesium chloride in the preparation of crystalline ceramic powders. Ceram Int 9:68–72
Kang J, Okabe TH (2013) Removal of iron from titanium ore through selective chlorination using magnesium chloride. Mater Trans 54:1444–1453
Huang J, Sohn I, Kang Y, Yang X (2022) Separation of Zn and Fe in ZnFe2O4 by reaction with MgCl2. Metall Mater Trans B 53:2634–2646
Wenz DA, Johnson I, Wolson RD (1969) CaCl2-rich region of the CaCl2-CaF2-CaO system. J Chem Eng Data 14:250–252
Guo T, Hu X, Matsuura H, Tsukihashi F, Zhou G (2010) Kinetics of Zn removal from ZnO–Fe2O3–CaCl2 System. ISIJ Int 50:1084–1088
Ito M, Morita K (2004) The solubility of MgO in molten MgCl2-CaCl2 salt. Mater Trans 45:2712–2718
Murugesan C, Okrasa L, Ugendar K, Chandrasekaran G, Liu X, Diao D, Shen J (2022) Improved magnetic and electrical properties of Zn substituted nanocrystalline MgFe2O4 ferrite. J Magn Magn Mater 550:169066
Da Silva S, Nakagomi F, Silva M, Franco A, Garg V, Oliveira A, Morais P (2012) Raman study of cations’ distribution in ZnxMg1−xFe2O4 nanoparticles. J Nanopart Res 14:1–10
Acknowledgements
The authors thank Dr. Yinjuan Chen and Mr. Ke Wang of Instrumentation and Service Center for Molecular Sciences at Westlake University for the assistance in the ICP-MS measurement and thank Ms. Ying Zhong and Dr. Xiaohe Miao of Instrumentation and Service Center for Physical Sciences at Westlake University for the XRD measurement.
Author information
Authors and Affiliations
Corresponding authors
Ethics declarations
Conflict of interest
On behalf of all authors, the corresponding author states that there is no conflict of interest.
Additional information
The contributing editor for this article was Hongmin Zhu.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
The original version of this article was revised: The graphic for Fig. 14 has been modified to fix a border between panels that was obscuring some of the information presented in the figure to correct a production error that was introduced into the article after the author reviewed proofs.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Huang, J., Li, G. & Yang, X. Chlorination of ZnFe2O4 by Molten MgCl2: Effect of Adding CaCl2. J. Sustain. Metall. 9, 1253–1266 (2023). https://doi.org/10.1007/s40831-023-00727-9
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s40831-023-00727-9