Recovery and purification of kish graphite from steelmaking dust by using acid leaching with physical auxiliary methods

Abstract

Graphite has been a critical raw material in recent years due to its high economic importance and high risk of supply disruptions. The traditional source of graphite is natural graphite ores; however, the production of natural graphite ores is controlled by few countries. It is thus important to find an alternative source of graphite, and steelmaking byproducts, such as dust and slag, can be considered because of their high potential for the recovery of kish graphite. The original kish graphite normally has a low carbon content and large amounts of impurities, and purification is necessary before using kish graphite in industry. The purpose of this study was to recover kish graphite from steelmaking dust by combining multi-stage froth flotation and acid leaching processes. The conditions of acid leaching were examined, and physical auxiliary methods (heating, microwave irradiation, and ultrasonication) were studied. After the multi-stage froth flotation process, the carbon content of the kish graphite was approximately 84 wt%, and Fe, Ca, Al, Na, K, Si, and Mg were the major elements of impurities. The acid leaching process was useful for enhancing the carbon content of the kish graphite and removing the impurities, particularly when using HCl and HBF₄. However, HCl should be a more appropriate selection for acid leaching when considering the price of acids. The carbon content of the kish graphite reached ~ 95 wt% when using 1.0 N HCl with ≥ 30 min of reaction time and a ≥ 5 L kg⁻¹ liquid-to-solid ratio. The physical auxiliary methods can further increase the carbon content of the kish graphite. The kish graphite purified by heating 1.0 N HCl at 80 °C for 5 min had the highest carbon content of approximately 97 wt%. The purified kish graphite and the natural graphite had similar crystallinity and lamellar structures, but the purified kish graphite had more structural defects. The recovery of kish graphite from steelmaking dust can obtain valuable materials and should have benefits for the environment.

1 Introduction

Graphite is an important industrial mineral due to its excellent physicochemical properties, e.g., electrical conductivity, thermal stability, lubrication, and chemical resistance, and it is thus widely used for battery electrodes, refractories, crucibles, castings, lubricants, foundry facings, brake linings, etc. The major source of graphite is natural graphite ores, including veins (lumps), flakes (crystalline), and amorphous materials [1]. Graphite is considered a critical raw material for both industry and national security because of its high economic importance and high risk of supply disruptions [2]. The search for alternative sources of graphite currently attracts increasing attention, and the recovery of kish graphite from steelmaking byproducts (e.g., dust and slag) is a possible option. The U.S. Geological Survey [3] reported that the recovery of high-quality kish graphite from steelmaking byproducts is technically feasible, but this process is currently not practiced due to the effects of the market price of graphite. The cost of recovering kish graphite is greater than that of mining natural graphite, but the difference is gradually narrowing when considering the market fluctuations and environmental benefits.

Kish graphite typically forms in steelmaking processes. In the Fe–C system, the carbon solubility decreases when molten iron cools [4], and an excessive amount of carbon precipitates as crystalline graphite in kish (a transitional layer between molten iron and slag). Liu and Loper [5] indicated that kish graphite has ideal or near-ideal crystalline structures, which are almost identical to those of natural graphite. However, kish graphite is quite different from natural graphite in terms of the presence of impurities. Chehreh Chelgani et al. [6] showed that gangue minerals, such as quartz, mica, feldspar, silicates, and carbonates, normally remain in natural graphite as impurities, whereas Jeon et al. [7] reported that minerals from steelmaking raw materials, including lime, iron oxides, silica, and alumina, usually exist in kish graphite. The purity, flake size, grade, and shape are highly related to the applications of graphite. For high-end technological applications, e.g., batteries, electronics, and composites, it is essential to remove impurities within graphite structures as much as possible. The carbon content of graphite should be at least 94–97 wt% for industrial use and must be higher than 99 wt% for electrical applications [1, 8]. The purity and flake size are also determinant factors for the price of graphite. Graphite with a higher purity and larger flake size certainly has a better price on the market [1, 3].

The purification methods for natural graphite typically include: acid–base methods, hydrofluoric acid (HF) methods, chlorination roasting methods, and high-temperature methods [1, 9,10,11,12,13]. The acid–base methods are the most widely used for the purification of natural graphite, and the principle is to use alkali fusion or solution (e.g., NaOH) to convert impurities (e.g., quartz and silicate) into soluble hydroxides and then remove the hydroxides with acid solutions [12]. HF methods are considered simple and rapid because HF can react with almost all impurities and then leaves water-soluble compounds that are subsequently washed off [10]. The acid–base and HF methods are effective and relatively inexpensive, but they are time-consuming and result in issues in the treatment of waste solutions. For chlorination roasting methods, graphite is roasted at a specific temperature (normally > 2000 °C) by introducing chlorine gas to chlorinate impurities. The impurities are thus converted into chloride compounds and then vaporized to the gas phase [11]. High-temperature methods are based on the principle of distillation, and in practice, graphite is heated under an inert atmosphere to a temperature (typically ~ 3000 °C) higher than the boiling point of impurities [13]. Although chlorination roasting and high-temperature methods can obtain high-purity graphite, they are very costly and carry some risks, such as occupational hazards and air pollution.

Due to the significant difference in impurities between natural graphite and kish graphite, the above purification methods may not be suitable for kish graphite. In addition, the purification methods for natural graphite have the disadvantages of consuming time or energy, potential environmental pollution, and high costs. The major impurities in kish graphite are normally lime and iron oxides, and an acid leaching method should be used to remove these impurities. Several previous studies have shown that physical methods, such as heating, microwave irradiation, and ultrasonication, are helpful for acid leaching [10, 14, 15]. Al-Sairafi et al. [10] used an HF solution to purify natural graphite and heated the solution to activate reactions. Li et al. [14] tried using microwave irradiation to assist acid leaching for the purification of natural graphite and found that the carbon content of graphite was increased by microwave irradiation. Vyas and Ting [15] reported that ultrasound has been applied to chemical leaching, including the leaching of metals and non-metal minerals, and it can usually promote reactions and improve leaching efficiency. Accordingly, the purpose of this study was to purify the kish graphite recovered from steelmaking dust by acid leaching with various types of acid solutions and to use physical auxiliary methods to improve the leaching efficiency. The similarity of the purified kish graphite to high-purity natural graphite was subsequently analyzed.

2 Materials and methods

2.1 Materials

In our previous study [16], kish graphite was beneficiated from steelmaking dust by using a multi-stage froth flotation process. The steelmaking dust was sampled in an integrated steel mill located in Kaohsiung, Taiwan. The multi-stage froth flotation process was conducted with a laboratory flotation machine (Metso, D12). The speed of the impeller was controlled at 1200 rpm, and the pulp density was set at 10%. The frother and collector used for froth flotation were methyl isobutyl carbinol (99%, Alfa Aesar) and kerosene (CPC Corporation, Taiwan), respectively. The multi-stage froth flotation process included three stages, and each stage had identical operating parameters. The kish graphite was progressively transferred to the concentrates stage by stage, while the impurities were moved to tailings. After the multi-stage froth flotation process, the carbon content of the kish graphite increased from approximately 34 to 84 wt%. This beneficiated kish graphite was employed as the raw material for the subsequent purification process.

2.2 Purification process of the kish graphite

To further increase the carbon content of the kish graphite, acid leaching methods were adopted for the purification process. Several types of acids, including hydrochloric acid (HCl, 37%, Sigma-Aldrich), nitric acid (HNO₃, 70%, Sigma-Aldrich), sulfuric acid (H₂SO₄, 95‒98%, Sigma-Aldrich) and tetrafluoroboric acid (HBF₄, 50%, Alfa Aesar), were used for the removal of impurities. Moreover, two types of mixed acids (MA1 and MA2) were prepared by mixing HCl and HBF₄ at ratios of 1:9 and 5:5, respectively. The concentrations of the acid solutions were between 0.1 and 1.0 N, and the liquid-to-solid (L/S) ratios were set at 5 and 10 L kg⁻¹. To promote the leaching of impurities, three physical auxiliary methods, i.e., heating, microwave irradiation, and ultrasonication, were tested. A digital hotplate stirrer (Heidolph, Hei-Plate Mix'n' Heat Core +) with a temperature sensor was used to heat the acid solutions to 60‒100 °C. The microwave irradiation was conducted with a microwave system (Milestone, Start D) under a temperature-controlled mode (40‒80 °C), and ultrasonication was performed by using a sonicator (Qsonica, Q700) with a standard 12.7-mm diameter probe. After a purification process, the liquid phase and kish graphite were separated with a vacuum filtration system, and the purified kish graphite was completely washed with deionized water for further analyses.

2.3 Analyses

As a comparison material, high-purity natural graphite (flake, 99.9 wt%, Thermo Scientific Chemicals) was employed to investigate the differences between natural graphite and purified kish graphite. Thermal analysis was conducted to measure the carbon content of the graphite samples by means of simultaneous differential scanning calorimetry and thermogravimetric analysis (DSC-TGA, TA Instrument, SDT 2960). Approximately 30 mg of sample was loaded into an alumina crucible, and the tube furnace temperature was programmed to ramp from ambient temperature to 1000 °C at a heating rate of 10 °C min⁻¹ under an air or nitrogen flow of 100 mL min⁻¹. The calculation of carbon content based on the DSC-TGA results was described in the previous study [16]. The chemical compositions of the impurities in the graphite samples were analyzed with an inductively coupled plasma-optical emission spectrometer (ICP-OES, PerkinElmer, Optima 2000 DV) following a microwave-assisted acid digestion procedure. The graphite samples were ground to fine powders less than 75 μm and digested using HNO₃, HCl, and HBF₄ at ~ 175 °C for 24 min with the microwave system, and the ICP-OES was used to determine the concentrations of elements in the digests. The chemical compositions of the impurities were calculated according to the digest volume, the elemental concentrations in digests, and the sample weight. An X-ray diffractometer (XRD, Bruker, D8 Advance) with Cu-Kα radiation was used to identify the mineralogical compositions of the graphite samples. To examine the structures and defects of the purified kish graphite and the natural graphite, Raman spectroscopy was conducted with a Horiba Jobin Yvon, LabRAM HR system. A high-resolution scanning electron microscope (SEM, Hitachi, SU8000) equipped with an energy dispersive spectrometer (EDS, Bruker, Quantax) was used to observe the morphology of the purified kish graphite and to analyze the elemental composition at specific positions of the surface. A conductive copper tape was used to attach samples to the SEM stage, thus reducing the interference in the EDS analysis for carbon.

3 Results and discussion

3.1 Purification of kish graphite with acid leaching

The kish graphite recovered from steelmaking dust was beneficiated by a multi-stage froth flotation process and then used as the raw material in this study. The chemical composition of the beneficiated kish graphite is given in Table 1. The carbon content reached over 84 wt%, a level that passes the lowest product grade of flake graphite (75 wt% [17]. In terms of impurities, small amounts of alkali metals (2.95 wt% of Na and 0.70 wt% of K), alkaline earth metals (0.32 wt% of Mg and 1.07 wt% of Ca), Al (1.90 wt%), and Si (0.55 wt%) were found in the beneficiated kish graphite. In addition, Fe-containing compounds should be the major impurities, and the amount of Fe in the beneficiated kish graphite reached 6.64 wt%. In our previous study, it was known that magnetite (Fe₃O₄) was the primary phase of the iron oxides present in the kish graphite [16]. The impurities found in this study were consistent with those reported by Jeon et al. [7], who indicated that during the formation of kish graphite, impurities (such as iron oxides, silica, alumina, magnesia, and lime) are inevitably encased in kish graphite. With regard to heavy metals, only trace amounts of heavy metals remained in the kish graphite after beneficiation, including Ti (492 mg kg⁻¹), Mn (413 mg kg⁻¹), and Zn (260 mg kg⁻¹). Toxic heavy metals of great concern, i.e., Cr, Cu, As, Se, Cd, Ba. Hg, and Pb, were all below detection limits.

Table 1 Chemical composition of the kish graphite before purification

To purify the beneficiated kish graphite for further grade improvement, different kinds of acids, including HCl, HNO₃, H₂SO₄, and HBF₄, were tested first. Although some studies have reported that HF is very effective at removing impurities from graphite [1, 6], especially silicates, HF was not used in this study because of safety and environmental concerns; thus, HBF₄ was adopted instead. To provide sufficient reaction conditions for acid leaching, the concentration of the acid solutions was 1.0 N, and the L/S ratio and reaction time were set at 10 L kg⁻¹ and 60 min, respectively. Figure 1 shows the carbon content of the purified kish graphite after acid leaching with various acid solutions. Using HCl for acid leaching afforded kish graphite with the highest carbon content (95.59 wt%), whereas using H₂SO₄ resulted in the lowest carbon content (92.50 wt%). The kish graphite prepared by using HCl for acid leaching had been repeated and analyzed many times, and the standard deviation and the relative standard deviation (RSD) were 0.43 wt% and 0.45%, respectively. Laverty et al. [18] indicated that using H₂SO₄ to purify graphite can cause the reprecipitation of gypsum (CaSO₄·2H₂O) on the surface of graphite when calcium is present. The carbon content of the kish graphite purified with HBF₄ was 94.46 wt%, which is close to that of the kish graphite purified with HCl.

Fig. 1

Carbon contents of the kish graphite after purification with various acid solutions

After the acid leaching process, the liquid phase was separated by means of vacuum filtering and then analyzed via the ICP-OES. Table 2 presents the leaching concentrations of metals in the purification process with various types of acids. The results showed that all the acids seemed useful for transferring the impurities from the kish graphite to the liquid phase. The differences in the leaching concentrations of Na, K, Mg, Al, Ti, Mn, and Zn among the HCl, HNO₃, H₂SO₄, and HBF₄ solutions were insignificant. When using HCl for acid leaching, Ca was effectively removed, and the leaching concentration of Ca was 1567 mg L⁻¹. HCl can also greatly remove Fe from kish graphite, and the leaching concentration of Fe reached 4599 mg L⁻¹. In terms of HNO₃, good removal of impurities from the kish graphite was observed, but the overall effect was lower than that of HCl. After the HNO₃ leaching process, the concentrations of Ca and Fe in the liquid phase were 1446 and 3540 mg L⁻¹, respectively. The effect of H₂SO₄ on the removal of impurities was lower than that of the other three acids, especially on Ca. The leaching concentration of Ca was only 529 mg L⁻¹, which is likely related to the reprecipitation of gypsum [18]. When HBF₄ was used for the acid leaching process, impurities were effectively removed, particularly Si and Fe. The concentration of Si in the liquid phase was 548.3 mg L⁻¹, which is approximately twice as high as that by using HCl, HNO₃ or H₂SO₄. The leaching concentration of Fe was also the highest (5245 mg L⁻¹). However, the Ca removal ability of HBF₄ was not outstanding, and the performance was slightly lower than that of HCl and HNO₃.

Table 2 Leaching concentrations of metals in the purification processes with various types of acids

Based on the above findings, it was suggested that using a mixed acid with HCl and HBF₄ may combine their advantages and thus enhance the removal of impurities from kish graphite. The carbon content of the kish graphite after the purification with mixed acids (MA1 and MA2) is also presented in Fig. 1, and the concentrations of metals in the liquid phases are listed in the last two columns of Table 2. When using the mixed acids, the carbon contents of the purified kish graphite were 95.46 wt% for both MA1 and MA2, which are comparable to those obtained by using HCl or HBF₄ individually. The mixed acids were also effective at transferring impurities to the liquid phase. High leaching concentrations of Ca, Si, and Fe were detected, but the improvements were not very significant. The concentrations of Ca, Si, and Fe were 1521, 562.5‒649.6, and 4710‒5122 mg L⁻¹, respectively. In comparison with MA1 and MA2, increasing the proportion of HBF₄ did not greatly improve the removal of impurities. Because the cost of HBF₄ is much greater than that of HCl, this study suggested that using HCl individually is more appropriate for the purification of kish graphite.

Figure 2 shows the carbon content of the kish graphite after purification with HCl at various L/S ratios and reaction times as a function of HCl strength. It was obvious that the carbon content of the kish graphite increased with increasing HCl concentration. The reaction time also affected the carbon content, but the increase was less significant after reacting for 30 min and later. When the reaction time was 30 min, purification at an L/S ratio of 5 L kg⁻¹ resulted in lower carbon contents of the kish graphite; however, the carbon content was close to that at an L/S ratio of 10 L kg⁻¹ if the concentration of HCl was 0.5 N or higher. In general, although the carbon content of the kish graphite after the purification process has positive correlations with the concentration of HCl, reaction time, and L/S ratio, the conditions of 1.0 N HCl with a reaction time of ≥ 30 min and an L/S ratio of ≥ 5 L kg⁻¹ should be sufficient for the purification process. Moreover, it seems that the carbon content of the kish graphite purified under ambient conditions was limited to approximately 95 wt%.

Fig. 2

Carbon contents of the kish graphite after purification with HCl at various L/S ratios and reaction times

3.2 Effects of physical auxiliary methods on the purification process

To further increase the carbon content of the purified kish graphite, physical auxiliary methods, including heating, microwave irradiation, and ultrasonication, were employed in this study. Table 3 shows the carbon content of the kish graphite after purification with HCl by heating and microwave irradiation. At a lower concentration of HCl (0.3 N), heating from ambient temperature to higher temperatures slightly increased the carbon content of the purified kish graphite. The carbon content of the kish graphite increased to 94.22 wt% when the HCl solution was heated to 100 °C. When using the microwave irradiation, the temperature-controlled mode was adopted, and the reaction temperatures were set at 40, 60, and 80 °C. At 40 and 60 °C, there was no improvement in the carbon content of the kish graphite. However, the carbon content of the kish graphite purified at 80 °C increased to 94.84 wt%. When using 1.0 N HCl for purification, heating the reaction temperature to 80 or 100 °C afforded kish graphite with 97.06 wt% of carbon content. In addition, it was found that the microwave irradiation slightly improved the purification process with 1.0 N HCl. The carbon content of the purified kish graphite increased to 96.15‒96.54 wt%. This study repeatedly prepared the kish graphite by heating at 80 °C for 30 min, and the standard deviation and the RSD were 0.38 wt% and 0.39%, respectively. To confirm that the experimental results are statistically significant, a right-tailed test was conducted with a hypothesis that the carbon content of kish graphite prepared by heating at 80 °C for 30 min is higher than that prepared at ambient temperature for 60 min. The calculated t value was 6.97 and the p value was much lower than 0.05, which means that these results have a significant difference.

Table 3 Carbon contents of the kish graphite after purification with HCl by using heating and microwave irradiation

This study also used ultrasonication as a physical auxiliary method for the purification process, and the amplitude of the sonicator was set at 25, 50, and 75%. Table 4 presents the carbon content of the kish graphite after purification with 1.0 N HCl by ultrasonication. The sonicator used in this study is capable of recording the input energy during the purification process, and it is known that the input energy increases with amplitude and reaction time. The temperature of the acid solution can also increase after absorbing the input energy. At a reaction time of 5 min, the temperature greatly increased with amplitude (from ambient temperature to 69 °C), but there was only a slight increase in the carbon content of the kish graphite. When the amplitude of the sonicator was set at 75%, the carbon content of the kish graphite increased with reaction time; however, the increase became insignificant at 5 min and later. Although some studies have indicated that ultrasonication is helpful for removing impurities from a material [15], the results of this research show that this improvement seems insignificant. This is probably attributed to the fact that some of the impurities in the kish graphite may be incorporated into the graphite structures.

Table 4 Carbon contents of the kish graphite after purification with HCl by using ultrasonication

From the above results of the purification of kish graphite, some appropriate processes were identified and the chemical compositions of the purified kish graphite samples were analyzed. The analysis results are given in Table 5. When purifying the kish graphite under ambient conditions, the carbon content reached 95.59 wt% after reacting with 1.0 N HCl at 10 L kg⁻¹ of L/S ratio for 60 min. A previous study obtained similar carbon content of the purified kish graphite (95.15 wt%) by combining water washing, magnetic separation, and acid leaching, but the concentrated HCl (37%) was used several times for acid leaching [19]. The major impurities remaining in the purified kish graphite (in order of concentration) were Fe (19,880 mg kg⁻¹), Al (10,720 mg kg⁻¹), Na (10,720 mg kg⁻¹), K (3876 mg kg⁻¹), and Si (1801 mg kg⁻¹). The other residual impurities, including Mg, Ca, and Ti, were below 1000 mg kg⁻¹. Three purification processes with physical auxiliary methods were suggested in this study, namely, heating at 80 °C for 30 min, microwave irradiation at 40 °C for 30 min, and ultrasonication at 75% amplitude for 5 min. Heating and microwave irradiation can further increase the carbon content but ultrasonication was not helpful. This should be due to the difference in reaction time. The optimal reaction time of ultrasonication concluded from Table 4 was 5 min, while that without any auxiliary treatment was 30 min. Although increasing the reaction time of ultrasonication from 5 to 30 min can slightly increase the carbon content of the kish graphite, the graphite flakes could be broken into small pieces due to excessive ultrasonication, which is not a desired result. Therefore, this study suggested that a reaction time of 5 min was relatively optimal for the ultrasonication treatment, and the carbon content was approximately limited to 94 wt%. It was also found that physical auxiliary methods were useful for reducing the residual impurities of kish graphite. Compared with those in the purification process under ambient conditions, the concentrations of Na, K, Mg, Ca, Al, and Fe in the purified kish graphite significantly decreased when the physical auxiliary methods were used. The concentration of Na markedly decreased from more than 10,000 to less than 500 mg kg⁻¹, and the residual Al decreased to approximately 3000 mg kg⁻¹. In terms of Fe, heating the HCl solution to 80 °C greatly reduced the residual Fe from 19,880 to 7787 mg kg⁻¹, whereas microwave irradiation and ultrasonication were not very helpful for Fe removal. However, it was noted that there were no significant effects on the residual concentrations of Si and Ti. In general, heating should be the easiest and most useful method for facilitating the purification of kish graphite. Microwave irradiation and ultrasonication can reduce the reaction temperature or time, but they have limitations in the removal of some impurities.

Table 5 Chemical composition of the kish graphite after purification with HCl by using physical auxiliary methods

3.3 Characterization of purified kish graphite

By means of a heating-assisted purification process, this study obtained purified kish graphite with 97.06 wt.% of carbon content, and its characteristics were subsequently examined. Figure 3 shows the TG and DSC curves of the purified kish graphite and the natural graphite under an air atmosphere. The TG curves exhibited a great weight loss, which should be related to the combustion of carbon under oxygen-containing conditions. Moreover, the DSC curves showed a significant exothermic peak at 550–950 °C, which is attributed to the release of energy from carbon combustion. Guo et al. [20] indicated that the oxidation of graphite at 660–675 °C was relatively slow, while rapid reactions of graphite oxidation occurred at 800–850 °C. This study revealed similar reactions related to graphite oxidation. The carbon content of the natural graphite analyzed in this study was 99.90 wt.%, which is consistent with the purity certificate provided by the manufacturer. Figure 3 shows that the TG and DSC curves of the purified kish graphite and the natural graphite were similar, and no reactions were observed for the other minerals in the thermal analysis.

Fig. 3

TG and DSC curves of purified kish graphite and natural graphite

Figure 4 presents the XRD patterns of the purified kish graphite and the natural graphite. The XRD pattern of the purified kish graphite was very similar to that of the natural graphite. In addition to the crystalline phase of graphite, no other minerals were present in the XRD patterns. The purified kish graphite and the natural graphite also had comparable diffraction intensities, which means that they have similar levels of crystallinity. The phenomenon of a preferred orientation was also observed in the XRD patterns. Theoretically, the strongest diffraction peak at ~ 26.5°2θ corresponds to the (0 0 2) crystal plane of graphite, and the second strongest diffraction peak (near 44.3°2θ) should correspond to the (1 0 1) crystal plane. However, in Fig. 4, two diffraction peaks corresponding to the (0 0 2) and (0 0 4) crystal planes were clearly observed near 26.5°2θ and 54.7°2θ, respectively. The intensities of these two peaks were greater than the theoretical peak intensities of graphite, which is attributed to the preferred orientation of graphite lamellae [21]. The above results indicate that the purified kish graphite is very similar to natural graphite; both have high crystallinity and lamellar structures.

Fig. 4

XRD patterns of purified kish graphite and natural graphite

Figure 5 shows the Raman spectra of the purified kish graphite and the natural graphite. Three clear peaks were observed in the Raman spectra. The first peak at approximately 1350 cm⁻¹ is called the D band, and it is absent in a well-ordered (defect-free) sample. The second peak at about 1582 cm⁻¹ is the G band, which is the most prominent feature in the Raman spectrum of graphitic materials. The third peak, which is located near 2700 cm⁻¹, is known as the 2D band. This band is also referred to as the G' band or the G* band in some studies [19, 22,23,24]. In Fig. 5, both the purified kish graphite and the natural graphite had distinct G and 2D bands; however, the D band was indistinct in the Raman spectrum of the natural graphite. This implies that natural graphite should have a nearly perfect structure, while purified kish graphite has some defects in its structure. The integrated intensity ratio for the D band and G band (I_D/I_G) has been used to evaluate the quantity of defects in graphitic materials [24]. The I_D/I_G value of the purified kish graphite was greater than that of the natural graphite, a result which proved that the purified kish graphite had more defects. These defects may be attributed to impurities that are incorporated into the structure of the kish graphite during the steelmaking process.

Fig. 5

Raman spectra of purified kish graphite and natural graphite

The SEM images of the purified kish graphite are presented in Fig. 6, and the EDS analysis results are given in Table 6. After the purification process, some impurity particles were still attached to the surface of the purified kish graphite. EDS analysis of an area (Position 1) indicated that the carbon content was 97.88%, and trace amounts of Mg, Al, Si, and Fe were detected. By analyzing the specific positions (Positions 2–6) with EDS, the results showed that the impurity particles were mainly composed of Fe, Al, and O. Accordingly, the possible minerals in the impurities were ferrite, alumina, and iron-aluminum oxides. To remove iron-containing particles, a secondary acid leaching procedure with oxalic acid may be considered in future work [25, 26]. In the SEM images, some cavities were also found on the surface of the purified kish graphite. These cavities should be created after removing the impurities that are incorporated in the graphite structures. The cavities can be beneficial for some applications of kish graphite. Krawczyk et al. [27] used kish graphite flakes as a cathode material for an aluminum chloride-graphite battery and noted that the cavities (holes) of kish graphite can facilitate the penetration of ionic liquids within the cathode. This enabled a second insertion pathway for AlCl₄⁻ ions and thus increased the energy density of the batteries.

Fig. 6

SEM images of the purified kish graphite

Table 6 EDS analysis results of the purified kish graphite

Figure 7 shows the flake size distribution and the carbon content in different flake size ranges of the purified kish graphite. The flake size ranges of 0.075–0.15 mm, 0.15–0.3 mm, and 0.3–0.6 mm had similar weight percentages between 22.5 and 26.8 wt%. Furthermore, they all had high carbon contents of 97.16–97.65 wt%. The flake size range of 0.6–1.18 mm also had a carbon content of 96.67 wt%, and its weight percentage was 10.6 wt%. A flake size larger than 1.18 mm accounted for less than 0.8 wt% and should thus be ignored. Their carbon contents were not available (NA) because the amount of sample was too less for analysis. On the other hand, flakes smaller than 0.075 mm had a carbon content of only 88.65 wt%, which is much lower than those in other flake size ranges. This finding showed that the impurities mostly remained in the flake size range of < 0.075 mm after the purification process. The price of graphite usually decreases with decreasing carbon content and flake size; therefore, the smallest flake size range (< 0.075 mm) may be removed from purified kish graphite due to its high impurity content and low market price. A similar strategy for the treatment of natural graphite ore was suggested in previous research [28].

Fig. 7

Distribution of flake size and carbon content in the purified kish graphite

In this study, the value of purified kish graphite was evaluated based on the carbon content and flake size. The U.S. Geological Survey [29] reported that the price of flake graphite (average value of imports at foreign ports) in the past three years was between 1300 and 1390 USD t⁻¹, and the price of flake graphite depends on the carbon content and flake size. The required carbon content for flake graphite products is 94‒97 wt%, and graphite with a larger flake size has a higher price on the market. The price of flake graphite product with a size of > 0.18 mm (80 mesh) was 1395 USD t⁻¹, while that with a size of < 0.15 mm (100 mesh) was only 755 USD t⁻¹ [30]. According to the relationship between price and flake size, the evaluated values of purified kish graphite recovered from steelmaking dust are given in Table 7. The recovery methods for kish graphite included the multi-stage froth flotation process developed in our previous study [16] and the purification process in this research. From one ton of steelmaking dust, 0.17 t of > 0.18 mm kish graphite can be obtained by means of the recovery method, and its value is estimated to be 236 USD. The recovery methods also yielded kish graphite with a 0.15–0.18 mm flake size (0.02 t, estimated value of 25 USD) and a < 0.15 mm flake size (0.14 t, estimated value of 106 USD). Accordingly, the total value of kish graphite recovered from one ton of steelmaking dust was estimated to be 367 USD. In contrast, the cost of treatment and disposal of steelmaking dust in Taiwan is approximately 700–1200 USD t⁻¹. This implies that replacing steelmaking dust treatment and disposal with a kish graphite recovery process has great potential benefit. Furthermore, the environmental risks of steelmaking dust treatment and disposal should be eliminated through practical implementation of kish graphite recovery. However, the current results of this study were not sufficient to establish a practical estimation of economic benefit. The investment evaluation of a real plant involves many parameters, such as treatment capacity and land acquisition, and the operating costs should also be assessed from the conditions of a real plant, including material and energy uses, waste and wastewater treatments, labor requirements, etc. These are far beyond the scope of this study and should be explored in depth separately.

Table 7 The value of purified kish graphite recovered from steelmaking dust

4 Conclusions

The following conclusions can be drawn from the present findings of this study. The carbon content of the kish graphite before purification was ~ 84 wt%, and the impurities were mainly composed of Fe, Ca, Al, Na, K, Si, and Mg, in order of concentration. Acid leaching methods were useful for enhancing the carbon content of kish graphite, especially when using HCl and HBF₄. HBF₄ solutions were capable of significantly removing Si, but this study suggested that HCl should be a more appropriate selection for acid leaching due to its much lower price. Using 1.0 N HCl with a reaction time of ≥ 30 min and an L/S ratio of ≥ 5 L kg⁻¹ should be sufficient for the purification of kish graphite, but the highest carbon content was limited to approximately 95 wt% under ambient conditions. The physical auxiliary methods, i.e., heating, microwave irradiation, and ultrasonication, were helpful for removing impurities from kish graphite. The use of microwave irradiation and ultrasonication increased the efficiency of acid leaching, and the kish graphite purified by heating 1.0 N HCl at 80 °C for 5 min had the highest carbon content of 97.06 wt%. The purified kish graphite and natural graphite had similar crystallinities and lamellar structures; however, the purified kish graphite had more defects in its structure. After the purification process, some cavities and residual impurity particles containing Fe, Al, and O were observed on the surface of the purified kish graphite. The total value of the kish graphite recovered from one ton of steelmaking dust was estimated to be 367 USD, and considering the cost of treatments and disposal for steelmaking dust, the recovery of kish graphite should have great potential benefits. To further increase the carbon content of kish graphite, a secondary acid leaching process may be investigated for the removal of Fe, and oxalic acid could be a candidate acid in future work. A full economic benefit assessment for the application of this technology is also recommended as needed in the future.