- Email: [email protected]
Recovery of iron and manganese from iron-bearing manganese residues by multi-step roasting and magnetic separation
Minerals Engineering 126 (2018) 177–183
Contents lists available at ScienceDirect
Minerals Engineering journal homepage: www.elsevier.com/locate/mineng
Recovery of iron and manganese from iron-bearing manganese residues by multi-step roasting and magnetic separation
T
⁎
Ning Penga,b,c, Qinglin Pana, Hui Liub,c, , Zhihui Yangb,c, Gongliang Wangb a
School of Material Science and Engineering, Central South University, 410083 Changsha, Hunan, China Institute of Environmental Science and Engineering, School of Metallurgy and Environment, Central South University, 410083 Changsha, Hunan, China c Chinese National Engineering Research Center for Control & Treatment of Heavy Metal Pollution, 410083 Changsha, Hunan, China b
A R T I C LE I N FO
A B S T R A C T
Keywords: Manganese residues Manganese ferrite Oxidative roasting Magnetic separation
Millions of tons of iron-bearing manganese residue are produced as a by-product of the electrolytic manganese industry. And the environmental contamination caused by manganese residue has received increasing attention. This paper focuses on the recovery of iron and manganese from high iron-bearing manganese residues. Manganese ferrite in manganese residues is initially decomposed by oxidative roasting, and the intermediates are magnetized in a reductive roasting step. The roasted product is milled and subjected to multi-stage magnetic separation. The optimum conditions are as follows: roasting at 750 °C under air flow for 30 min, roasting in a CO atmosphere at 750 °C for 30 min, and separating under a magnetic intensity of 1000 G for weak magnetic separation and of 12,000 G for strong magnetic separation. The recovery and grade of iron in the iron concentrate were 72.29% and 62.21%, respectively, and those of manganese in the manganese concentrate were 90.75% and 35.21%, respectively. This study demonstrates that the combination of roasting and magnetic separation provides a promising process for the recovery of iron and manganese from high-iron-bearing manganese residues.
1. Introduction
leaching (Duan et al., 2011; Xin et al., 2011) have all been applied, with the maximum manganese extraction achieved in the bio-leaching process at above 90%. However, traditional chemical leaching processes have disadvantages, including massive consumption of solvent and generation of secondary high iron-bearing residues in the iron precipitation process. Bioleaching would the most promising method if the leaching kinetics could be accelerated, as the leaching time of this method was counted in days (Duan et al., 2011). In this study, roasting and magnetic separation techniques were applied for the recycling of iron and manganese from iron-bearing manganese residues. Manganese ferrite was first decomposed into magnetite and low-valence manganese oxide using a two-stage roasting process. The roasted product was then subjected to ball-milling and multi-stage magnetic separation to recover manganese and iron separately. The decomposition mechanism of manganese ferrite was analysed by thermodynamic calculation and XRD analysis of the roasted product, and the operating parameters of the roasting and magnetic separation procedures were studied.
Electrolytic manganese residues (EMR) are produced in significant amounts from the electrolytic manganese metal (EMM) industry, with 6–9 tons of residue generated per ton of metal (Duan et al., 2010). Most of these residues remain untreated and are generally stockpiled. The hazardous metal elements (Mn, Zn, Cu, Pb, Cd, Cr) (Zhou et al., 2014) associated with these residues result in high environmental risks and waste potentially recoverable valuable resources (Du et al., 2015; Li et al., 2014; Yan and Qiu, 2014). The recycling and recovery of manganese from EMR has aroused increasing attention in recent years. Manganese is mainly recovered by the hydro-metallurgical route using a hot acid leaching (HAL) process with sulfuric acid as a solvent. The maximum manganese extraction using this process varies because manganese is only soluble in its lowvalence form (Elsherief, 2000; Liu et al., 2014). In addition to the HAL process, citric acid leaching assisted by ultrasound (Li et al., 2008; Yuzhu et al., 2007), sulfuric acid leaching with glucose and saccharose (Yao et al., 2003), organic acid leaching (Das et al., 2012) and bio-
⁎ Corresponding author at: Institute of Environmental Science and Engineering, School of Metallurgy and Environment, Central South University, 410083 Changsha, Hunan, China. E-mail address: [email protected] (H. Liu).
https://doi.org/10.1016/j.mineng.2018.07.002 Received 21 May 2018; Received in revised form 8 July 2018; Accepted 10 July 2018 0892-6875/ © 2018 Elsevier Ltd. All rights reserved.
Minerals Engineering 126 (2018) 177–183
N. Peng et al.
introduced when the desired temperature was reached. After a certain roasting duration, the reaction gas was replaced by argon, and then the roasting product was quenched in water. The roasting temperature ranged from 600 °C to 850 °C at 50 °C increments, and roasting times ranging from 10 min to 40 min at 5 min increments were examined in the oxidative roasting experiments. In the reductive roasting experiments, the roasting temperature, roasting time and concentration of CO were maintained in the ranges of 600–800 °C, 10–40 min and 2–12%, respectively. The material recovered after quenching was directly subjected to mechanical milling and magnetic separation.
Table 1 Element content in EMR (wt%). Elements
Content
Elements
Content
O Si Al S Ca K Cl Mg
35.01 18.63 3.08 2.32 1.56 1.41 0.02 0.39
Mn Fe P Cu Zn Ni Cr Cd
15.12 16.79 0.734 0.05 0.24 0.01 – –
2.5. Mechanical milling and magnetic separation Mechanical milling was performed in a planetary ball mill operating at 500 rpm. Approximately 200 g of steel balls were kept in each cell, along with 20 g of the sample powder. Magnetic separation experiments were carried out with a Kolm-type high-gradient magnetic separator (Oberteuffer, 1974). With this separator, the slurry obtained after mechanical milling was placed into a plastic container equipped with a stirrer and diluted using tap water. Then, the prepared fluid was introduced into the separation chamber under gravity, in which a magnetic field perpendicular to the direction of flow was generated using an electromagnet. The magnetic intensity ranged from 800 to 1400 G for iron separation and from 8000 to 12,000 G for manganese separation. Magnetic particles were enriched on the plates as the concentrate, and non-magnetic particles were collected as the tailings. After separation, the concentrate was washed, filtered and dried for further analyses.
Fig. 1. XRD pattern of EMR.
2. Material and methods 2.1. Materials A 100 kg sample of EMR was obtained from an electrolytic manganese plant located in Hunan, China. The sample was sieved to produce a fraction of below 75 μm size and dried at 105 °C to a constant weight. ICP analyses of the EMR are shown in Table 1. Its corresponding XRD pattern is shown in Fig. 1. It can be seen Fig. 1 that the iron was associated with manganese in the residues, with MnFe2O4 was determined as the main Mn-bearing component in the EMR. 2.2. Analysis Potassium dichromate titration of Fe and Mn was used for chemical analysis of the samples. XRD analysis with Cu Kα-radiation (Rigaku, TTR-III) was applied for analysis of the primary residue and for examining the phase transformations during roasting. The multiple element content was determined by ICP-AES analysis (Baird, PS-6). 2.3. Thermodynamic analysis The thermodynamic analysis was carried out using FactSage (Källén et al., 2014; Sorensen et al., 2010). The predominance area diagram of MnFe2O4-O2 and MnF2O4-CO in the range 100–1000 °C were calculated based on minimization of the Gibbs free energy. 2.4. Roasting Fig. 2. Predominance area diagram of MnFe2O4-O2 (A), MnFe2O4-CO (B), the reddish region is the unstable area for MnFe2O4. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
The roasting process was conducted using a pipe furnace (Peng et al., 2017). Crucibles with a volume of 300 ml were plastered with a layer of sample (20 g, 2–3 mm in thickness) on the walls and roasted under a flowing stream of reaction gas (CO + Ar or air). The gas was 178
Minerals Engineering 126 (2018) 177–183
N. Peng et al.
Fig. 3. Effect of temperature (A), roasting time (B) on the decomposition of manganese ferrite and the XRD patterns of the roasted products at various temperatures (C).
3. Results and discussion
10 min to 40 min at 5 min increments were examined. The results are shown in Fig. 3. In Fig. 3(A), the decomposition ratio χ reached its maximum at 750 °C and then decreased with further increase of the temperature. In Fig. 3(B), the decomposition ratio increased with roasting time and reached a maximum after roasting at 750 °C for 30 min, indicating that 30 min was sufficient for efficient decomposition. The XRD patterns of the roasted product at the different temperatures are shown in Fig. 3(C). The height of the characteristic peaks of manganese ferrite was lowered with increasing temperature. In contrast, the characteristic peaks of haematite, manganic oxide and manganous oxide increased. This result indicated the breakdown of the crystal structure of manganese ferrite. However, the characteristic peak height of manganese ferrite slightly increased at 850 °C ascribing to the reform of manganese ferrite, which could be explained by the phase transformation from the 7th phase area to the 8th in Fig. 2(A) with the increasing of temperature. Meanwhile, (Fe, Mn)2SiO4 was formed at 850 °C because of the abundant association of silicon dioxide. Consequently, the decomposition ratio of manganese ferrite decreased slightly in Fig. 3(A). The optimal conditions indicate that the ore should be roasted at 750 °C for 30 min. The particle size distribution of the samples before and after roasting at 750 °C for 30 min (see Fig. 4) showed an obvious increase in the grain size. A milling pre-treatment may be needed for efficient separation of the iron and manganese.
3.1. Experimental mechanism focusing on the selective decomposition of manganese ferrite Fig. 2 shows the calculated phase diagrams of the binary systems MnFe2O4-O2 and MnFe2O4-CO. In Fig. 2(A), MnFe2O4 became unstable with increasing oxygen molar ratio and decreasing temperature. The upper limit temperature of the unstable region of MnFe2O4 was 815 °C, and the left limit oxygen molar ratio was 0.2, in which the area MnFe2O4 was completely decomposed into Fe2O3 and manganese oxides. As seen in Fig. 2(B), the predominant phases were monoxide solid solution ((Mn2+, Fe2+)O) and spinel solid solution ((Fe2+, Fe3+, Mn2+)3O4). The spinel phase was not stable at CO molar ratios > 0.58 and temperatures > 613 °C; however, the monoxide solid solution still existed in this region. The formation of the solid solution may be disadvantageous to the separation of iron from manganese.
3.2. Effect of oxidative roasting conditions on the decomposition of manganese ferrite The EMR was roasted in air flow with a flow rate of 1 L/min in a series of oxidative roasting experiments. Roasting temperatures ranging from 600 °C to 850 °C at 50 °C increments and times ranging from
Fig. 4. Particle size distribution of EMR before and after roasting at 750 °C for 30 min.
179
Minerals Engineering 126 (2018) 177–183
N. Peng et al.
Fig. 5. Flow sheet of process.
reduction of haematite was verified through the variation of the characteristic peaks of magnetite and haematite. In Section 3.1, we can deduce from Fig. 2(B) that manganese ferrite is unstable only in high CO concentration temperature (in the 7th-11th phase region). However, the reform of manganese ferrite in weak reductive roasting is negligible as the characteristic peaks of manganese ferrite are undetected in Fig. 7. This may be caused by the generation of magnetite by reduction of haematite, for the form of iron is trivalence in MnFe2O4. Meanwhile, the height of the diffraction peaks of Mn3O4 was elevated with increasing roasting temperature, which was due to the partial decomposition of Mn2O3. This transformation may be in favour of the separation of manganese, since the manganese recovery and grade in Fig. 6(B) increased above 750 °C.
3.3. Effect of reductive roasting conditions on the magnetic separation of iron and manganese After oxidative roasting, the manganese ferrite was decomposed into haematite, manganic oxide and manganous oxide. To separate the iron from the manganese, magnetic roasting of the roasted product of oxidative roasting was then employed to transform the haematite into magnetite. The magnetite was recovered by magnetic separation, and the manganese oxides were recovered by strong magnetic separation. The process flow sheet was designed accordingly and is shown in Fig. 5, wherein the magnetic field intensities of the weak and strong magnetic separation stages were 1000 G and 12,000 G, respectively. In the reductive roasting experiments, a gas mixture of carbon monoxide and argon was used as a reductant, and the influence of the roasting temperature, time and concentration of carbon monoxide on the magnetic separation of iron and manganese was examined. It can be seen from Fig. 6(A) and (B) higher temperatures gave greater iron and manganese recovery below 750 °C, but the recovery and grade of iron decreased with further increase in temperature due to over-reduction of the iron oxides. Fig. 6(C) and (D) illustrates the time dependence of metal separation. The roasting time was significant for iron recovery because both the iron recovery and grade decrease after 30 min. The roasting time of 30 min is, therefore, sufficient for the separation of manganese. The effect of the carbon monoxide concentration on the metal separation is shown in Fig. 6(F) and (G). The maximum iron recovery and grade occurred when the CO concentration was 10%. However, the recovery of manganese fluctuated with the CO concentration, and a CO concentration of 10% was considered suitable for manganese separation. The roasted products from the various tests were examined by X-ray diffraction, as shown in Fig. 7. The generation of magnetite by
3.4. Effect of milling time and magnetic intensity on the magnetic separation of iron and manganese The influence of milling time and magnetic intensity on the magnetic separation of iron and manganese was studied on the basis of the above test work. It can be seen from Fig. 8 (A) and (B) that the recovery and grade of iron reached a plateau after milling for 20 min; however, the recovery and grade of manganese decreased slightly after milling for 30 min because grinding could have altered the crystal structure of manganese oxide in the roasted product to yield an amorphous mixture (Bid and Pradhan, 2003). In Fig. 8(C) and (D), both iron and manganese increased with increasing magnetic intensity, while the grade of both changed conversely. To yield an acceptable iron and manganese separation, the optimal magnetic intensity for iron separation was 1000–1200 G and that for manganese 11,000–12,000 G. The XRD patterns for both iron and manganese concentrates in Fig. 8(E) and (F) show that the main phases in the iron concentrate were Fe3O4 and SiO2 180
Minerals Engineering 126 (2018) 177–183
N. Peng et al.
°C
°C
Fig. 6. Effect of temperature (A, B), roasting duration (C, D) and CO concentration (E, F) on iron and manganese magnetic separation.
and those of the manganese concentrate were SiO2, MnO and Mn3O4. The roasted product was subjected to multi-step magnetic separation under the optimal magnetic intensity, and the element balance of iron and manganese across the whole process is shown in Table 2. The grade of iron and manganese reached 62.21% and 35.21%, respectively. 4. Conclusions This study showed that the decomposition of manganese ferrite MnFe2O4 is important for the recycling of high-iron-bearing manganese residues. An oxidative atmosphere favours the decomposition of manganese ferrite, which was confirmed by thermodynamic calculation and experiments. Haematite was further reduced to magnetite in a reducing roast, which caused the generation of manganous oxide and manganous-manganic oxide. In this process, the regeneration of manganese ferrite through the reaction between manganous oxide and haematite was found to be negligible. Iron and manganese in the roasted product were dissociated by a milling pretreatment and then recovered in a two-
Fig. 7. XRD spectrums of the reductively roasted products at various temperature. 181
Minerals Engineering 126 (2018) 177–183
N. Peng et al.
Fig. 8. Effect of milling time (A, B) and magnetic intensity (C, D) on the magnetic separation of iron (A, C) and manganese (B, D) and XRD patterns of the iron (E) and manganese (F) concentrates.
Acknowledgements
Table 2 Iron and manganese balance of whole process (wt%). Samples
Grade
Yield
Fe
Mn
16.79 62.21 8.45 2.51
15.12 4.26 35.21 1.87
The authors would like to thank the Natural Science Foundation of China (51574295) for financial support for this study.
Recovery Fe
Mn
100 72.29 19.61 6.21
100 5.50 90.75 5.14
References EMR Iron concentration Manganese concentration Tailings
100 19.51 38.97 41.52
Bid, S., Pradhan, S.K., 2003. Preparation of zinc ferrite by high-energy ball-milling and microstructure characterization by Rietveld’s analysis. Mater. Chem. Phys. 82 (1), 27–37. Das, A.P., Swain, S., Panda, S., Pradhan, N., Sukla, L.B., 2012. Reductive acid leaching of low grade manganese ores. Geomaterials 2 (04), 70. Du, B., Hou, D., Duan, N., Zhou, C., Wang, J., Dan, Z., 2015. Immobilization of high concentrations of soluble Mn(II) from electrolytic manganese solid waste using inorganic chemicals. Environ. Sci. Pollut. Res. 22 (10), 7782–7793. Duan, N., Fan, W., Changbo, Z., Chunlei, Z., Hongbing, Y., 2010. Analysis of pollution materials generated from electrolytic manganese industries in China. Resour. Conserv. Recycl. 54 (8), 506–511. Duan, N., Zhou, C., Chen, B., Jiang, W., Xin, B., 2011. Bioleaching of Mn from manganese residues by the mixed culture of Acidithiobacillus and mechanism. J. Chem. Technol.
step magnetic separation. The enrichment of iron and manganese and the production of a low-valence manganese oxide make this process promising and attractive for the recycling of high-iron-bearing manganese residues.
182
Minerals Engineering 126 (2018) 177–183
N. Peng et al.
Ag and In from high iron-bearing zinc calcine. J. Central South Univ. 24 (5), 1082–1089. Sorensen, B., Gaal, S., Ringdalen, E., Tangstad, M., Kononov, R., Ostrovski, O., 2010. Phase compositions of manganese ores and their change in the process of calcination. Int. J. Miner. Process. 94 (3), 101–110. Xin, B., Chen, B., Duan, N., Zhou, C., 2011. Extraction of manganese from electrolytic manganese residue by bioleaching. Bioresour. Technol. 102 (2), 1683–1687. Yan, S., Qiu, Y.-R., 2014. Preparation of electronic grade manganese sulfate from leaching solution of ferromanganese slag. Trans. Nonferrous Metals Soc. Chin. 24 (11), 3716–3721. Yao, J., Cheng, S., Xiao, S.-H., 2003. Influence of lixiviate rate of manganese in rhodochrosite by adding addition agent. J.-Jishou Univ. Nat. Sci. 24 (1), 43–45. Yuzhu, O., Xiaowei, P., Jianbing, C., Zhiping, L., Xiaodong, D., 2007. Ultrasonic leaching of electrolytic manganese residue with additive. Environ. Protect. Chem. Ind. 27 (3), 257–259. Zhou, C., Du, B., Wang, N., Chen, Z., 2014. Preparation and strength property of autoclaved bricks from electrolytic manganese residue. J. Cleaner Prod. 84, 707–714.
Biotechnol. 86 (6), 832–837. Elsherief, A., 2000. A study of the electroleaching of manganese ore. Hydrometallurgy 55 (3), 311–326. Källén, M., Hallberg, P., Rydén, M., Mattisson, T., Lyngfelt, A., 2014. Combined oxides of iron, manganese and silica as oxygen carriers for chemical-looping combustion. Fuel Process. Technol. 124, 87–96. Li, C., Zhong, H., Wang, S., Xue, J., 2014. Leaching behavior and risk assessment of heavy metals in a landfill of electrolytic manganese residue in Western Hunan, China. Hum. Ecol. Risk Assess.: Int. J. 20 (5), 1249–1263. Li, H., Zhang, Z., Tang, S., Li, Y., Zhang, Y., 2008. Ultrasonically assisted acid extraction of manganese from slag. Ultrason. Sonochem. 15 (4), 339–343. Liu, Y., Lin, Q., Li, L., Fu, J., Zhu, Z., Wang, C., Qian, D., 2014. Study on hydrometallurgical process and kinetics of manganese extraction from low-grade manganese carbonate ores. Int. J. Min. Sci. Technol. 24 (4), 567–571. Oberteuffer, J., 1974. Magnetic separation: a review of principles, devices, and applications. IEEE Trans. Magn. 10 (2), 223–238. Peng, B., Peng, N., Liu, H., Xue, K., Lin, D.-H., 2017. Comprehensive recovery of Fe, Zn,
183
Contents lists available at ScienceDirect
Minerals Engineering journal homepage: www.elsevier.com/locate/mineng
Recovery of iron and manganese from iron-bearing manganese residues by multi-step roasting and magnetic separation
T
⁎
Ning Penga,b,c, Qinglin Pana, Hui Liub,c, , Zhihui Yangb,c, Gongliang Wangb a
School of Material Science and Engineering, Central South University, 410083 Changsha, Hunan, China Institute of Environmental Science and Engineering, School of Metallurgy and Environment, Central South University, 410083 Changsha, Hunan, China c Chinese National Engineering Research Center for Control & Treatment of Heavy Metal Pollution, 410083 Changsha, Hunan, China b
A R T I C LE I N FO
A B S T R A C T
Keywords: Manganese residues Manganese ferrite Oxidative roasting Magnetic separation
Millions of tons of iron-bearing manganese residue are produced as a by-product of the electrolytic manganese industry. And the environmental contamination caused by manganese residue has received increasing attention. This paper focuses on the recovery of iron and manganese from high iron-bearing manganese residues. Manganese ferrite in manganese residues is initially decomposed by oxidative roasting, and the intermediates are magnetized in a reductive roasting step. The roasted product is milled and subjected to multi-stage magnetic separation. The optimum conditions are as follows: roasting at 750 °C under air flow for 30 min, roasting in a CO atmosphere at 750 °C for 30 min, and separating under a magnetic intensity of 1000 G for weak magnetic separation and of 12,000 G for strong magnetic separation. The recovery and grade of iron in the iron concentrate were 72.29% and 62.21%, respectively, and those of manganese in the manganese concentrate were 90.75% and 35.21%, respectively. This study demonstrates that the combination of roasting and magnetic separation provides a promising process for the recovery of iron and manganese from high-iron-bearing manganese residues.
1. Introduction
leaching (Duan et al., 2011; Xin et al., 2011) have all been applied, with the maximum manganese extraction achieved in the bio-leaching process at above 90%. However, traditional chemical leaching processes have disadvantages, including massive consumption of solvent and generation of secondary high iron-bearing residues in the iron precipitation process. Bioleaching would the most promising method if the leaching kinetics could be accelerated, as the leaching time of this method was counted in days (Duan et al., 2011). In this study, roasting and magnetic separation techniques were applied for the recycling of iron and manganese from iron-bearing manganese residues. Manganese ferrite was first decomposed into magnetite and low-valence manganese oxide using a two-stage roasting process. The roasted product was then subjected to ball-milling and multi-stage magnetic separation to recover manganese and iron separately. The decomposition mechanism of manganese ferrite was analysed by thermodynamic calculation and XRD analysis of the roasted product, and the operating parameters of the roasting and magnetic separation procedures were studied.
Electrolytic manganese residues (EMR) are produced in significant amounts from the electrolytic manganese metal (EMM) industry, with 6–9 tons of residue generated per ton of metal (Duan et al., 2010). Most of these residues remain untreated and are generally stockpiled. The hazardous metal elements (Mn, Zn, Cu, Pb, Cd, Cr) (Zhou et al., 2014) associated with these residues result in high environmental risks and waste potentially recoverable valuable resources (Du et al., 2015; Li et al., 2014; Yan and Qiu, 2014). The recycling and recovery of manganese from EMR has aroused increasing attention in recent years. Manganese is mainly recovered by the hydro-metallurgical route using a hot acid leaching (HAL) process with sulfuric acid as a solvent. The maximum manganese extraction using this process varies because manganese is only soluble in its lowvalence form (Elsherief, 2000; Liu et al., 2014). In addition to the HAL process, citric acid leaching assisted by ultrasound (Li et al., 2008; Yuzhu et al., 2007), sulfuric acid leaching with glucose and saccharose (Yao et al., 2003), organic acid leaching (Das et al., 2012) and bio-
⁎ Corresponding author at: Institute of Environmental Science and Engineering, School of Metallurgy and Environment, Central South University, 410083 Changsha, Hunan, China. E-mail address: [email protected] (H. Liu).
https://doi.org/10.1016/j.mineng.2018.07.002 Received 21 May 2018; Received in revised form 8 July 2018; Accepted 10 July 2018 0892-6875/ © 2018 Elsevier Ltd. All rights reserved.
Minerals Engineering 126 (2018) 177–183
N. Peng et al.
introduced when the desired temperature was reached. After a certain roasting duration, the reaction gas was replaced by argon, and then the roasting product was quenched in water. The roasting temperature ranged from 600 °C to 850 °C at 50 °C increments, and roasting times ranging from 10 min to 40 min at 5 min increments were examined in the oxidative roasting experiments. In the reductive roasting experiments, the roasting temperature, roasting time and concentration of CO were maintained in the ranges of 600–800 °C, 10–40 min and 2–12%, respectively. The material recovered after quenching was directly subjected to mechanical milling and magnetic separation.
Table 1 Element content in EMR (wt%). Elements
Content
Elements
Content
O Si Al S Ca K Cl Mg
35.01 18.63 3.08 2.32 1.56 1.41 0.02 0.39
Mn Fe P Cu Zn Ni Cr Cd
15.12 16.79 0.734 0.05 0.24 0.01 – –
2.5. Mechanical milling and magnetic separation Mechanical milling was performed in a planetary ball mill operating at 500 rpm. Approximately 200 g of steel balls were kept in each cell, along with 20 g of the sample powder. Magnetic separation experiments were carried out with a Kolm-type high-gradient magnetic separator (Oberteuffer, 1974). With this separator, the slurry obtained after mechanical milling was placed into a plastic container equipped with a stirrer and diluted using tap water. Then, the prepared fluid was introduced into the separation chamber under gravity, in which a magnetic field perpendicular to the direction of flow was generated using an electromagnet. The magnetic intensity ranged from 800 to 1400 G for iron separation and from 8000 to 12,000 G for manganese separation. Magnetic particles were enriched on the plates as the concentrate, and non-magnetic particles were collected as the tailings. After separation, the concentrate was washed, filtered and dried for further analyses.
Fig. 1. XRD pattern of EMR.
2. Material and methods 2.1. Materials A 100 kg sample of EMR was obtained from an electrolytic manganese plant located in Hunan, China. The sample was sieved to produce a fraction of below 75 μm size and dried at 105 °C to a constant weight. ICP analyses of the EMR are shown in Table 1. Its corresponding XRD pattern is shown in Fig. 1. It can be seen Fig. 1 that the iron was associated with manganese in the residues, with MnFe2O4 was determined as the main Mn-bearing component in the EMR. 2.2. Analysis Potassium dichromate titration of Fe and Mn was used for chemical analysis of the samples. XRD analysis with Cu Kα-radiation (Rigaku, TTR-III) was applied for analysis of the primary residue and for examining the phase transformations during roasting. The multiple element content was determined by ICP-AES analysis (Baird, PS-6). 2.3. Thermodynamic analysis The thermodynamic analysis was carried out using FactSage (Källén et al., 2014; Sorensen et al., 2010). The predominance area diagram of MnFe2O4-O2 and MnF2O4-CO in the range 100–1000 °C were calculated based on minimization of the Gibbs free energy. 2.4. Roasting Fig. 2. Predominance area diagram of MnFe2O4-O2 (A), MnFe2O4-CO (B), the reddish region is the unstable area for MnFe2O4. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
The roasting process was conducted using a pipe furnace (Peng et al., 2017). Crucibles with a volume of 300 ml were plastered with a layer of sample (20 g, 2–3 mm in thickness) on the walls and roasted under a flowing stream of reaction gas (CO + Ar or air). The gas was 178
Minerals Engineering 126 (2018) 177–183
N. Peng et al.
Fig. 3. Effect of temperature (A), roasting time (B) on the decomposition of manganese ferrite and the XRD patterns of the roasted products at various temperatures (C).
3. Results and discussion
10 min to 40 min at 5 min increments were examined. The results are shown in Fig. 3. In Fig. 3(A), the decomposition ratio χ reached its maximum at 750 °C and then decreased with further increase of the temperature. In Fig. 3(B), the decomposition ratio increased with roasting time and reached a maximum after roasting at 750 °C for 30 min, indicating that 30 min was sufficient for efficient decomposition. The XRD patterns of the roasted product at the different temperatures are shown in Fig. 3(C). The height of the characteristic peaks of manganese ferrite was lowered with increasing temperature. In contrast, the characteristic peaks of haematite, manganic oxide and manganous oxide increased. This result indicated the breakdown of the crystal structure of manganese ferrite. However, the characteristic peak height of manganese ferrite slightly increased at 850 °C ascribing to the reform of manganese ferrite, which could be explained by the phase transformation from the 7th phase area to the 8th in Fig. 2(A) with the increasing of temperature. Meanwhile, (Fe, Mn)2SiO4 was formed at 850 °C because of the abundant association of silicon dioxide. Consequently, the decomposition ratio of manganese ferrite decreased slightly in Fig. 3(A). The optimal conditions indicate that the ore should be roasted at 750 °C for 30 min. The particle size distribution of the samples before and after roasting at 750 °C for 30 min (see Fig. 4) showed an obvious increase in the grain size. A milling pre-treatment may be needed for efficient separation of the iron and manganese.
3.1. Experimental mechanism focusing on the selective decomposition of manganese ferrite Fig. 2 shows the calculated phase diagrams of the binary systems MnFe2O4-O2 and MnFe2O4-CO. In Fig. 2(A), MnFe2O4 became unstable with increasing oxygen molar ratio and decreasing temperature. The upper limit temperature of the unstable region of MnFe2O4 was 815 °C, and the left limit oxygen molar ratio was 0.2, in which the area MnFe2O4 was completely decomposed into Fe2O3 and manganese oxides. As seen in Fig. 2(B), the predominant phases were monoxide solid solution ((Mn2+, Fe2+)O) and spinel solid solution ((Fe2+, Fe3+, Mn2+)3O4). The spinel phase was not stable at CO molar ratios > 0.58 and temperatures > 613 °C; however, the monoxide solid solution still existed in this region. The formation of the solid solution may be disadvantageous to the separation of iron from manganese.
3.2. Effect of oxidative roasting conditions on the decomposition of manganese ferrite The EMR was roasted in air flow with a flow rate of 1 L/min in a series of oxidative roasting experiments. Roasting temperatures ranging from 600 °C to 850 °C at 50 °C increments and times ranging from
Fig. 4. Particle size distribution of EMR before and after roasting at 750 °C for 30 min.
179
Minerals Engineering 126 (2018) 177–183
N. Peng et al.
Fig. 5. Flow sheet of process.
reduction of haematite was verified through the variation of the characteristic peaks of magnetite and haematite. In Section 3.1, we can deduce from Fig. 2(B) that manganese ferrite is unstable only in high CO concentration temperature (in the 7th-11th phase region). However, the reform of manganese ferrite in weak reductive roasting is negligible as the characteristic peaks of manganese ferrite are undetected in Fig. 7. This may be caused by the generation of magnetite by reduction of haematite, for the form of iron is trivalence in MnFe2O4. Meanwhile, the height of the diffraction peaks of Mn3O4 was elevated with increasing roasting temperature, which was due to the partial decomposition of Mn2O3. This transformation may be in favour of the separation of manganese, since the manganese recovery and grade in Fig. 6(B) increased above 750 °C.
3.3. Effect of reductive roasting conditions on the magnetic separation of iron and manganese After oxidative roasting, the manganese ferrite was decomposed into haematite, manganic oxide and manganous oxide. To separate the iron from the manganese, magnetic roasting of the roasted product of oxidative roasting was then employed to transform the haematite into magnetite. The magnetite was recovered by magnetic separation, and the manganese oxides were recovered by strong magnetic separation. The process flow sheet was designed accordingly and is shown in Fig. 5, wherein the magnetic field intensities of the weak and strong magnetic separation stages were 1000 G and 12,000 G, respectively. In the reductive roasting experiments, a gas mixture of carbon monoxide and argon was used as a reductant, and the influence of the roasting temperature, time and concentration of carbon monoxide on the magnetic separation of iron and manganese was examined. It can be seen from Fig. 6(A) and (B) higher temperatures gave greater iron and manganese recovery below 750 °C, but the recovery and grade of iron decreased with further increase in temperature due to over-reduction of the iron oxides. Fig. 6(C) and (D) illustrates the time dependence of metal separation. The roasting time was significant for iron recovery because both the iron recovery and grade decrease after 30 min. The roasting time of 30 min is, therefore, sufficient for the separation of manganese. The effect of the carbon monoxide concentration on the metal separation is shown in Fig. 6(F) and (G). The maximum iron recovery and grade occurred when the CO concentration was 10%. However, the recovery of manganese fluctuated with the CO concentration, and a CO concentration of 10% was considered suitable for manganese separation. The roasted products from the various tests were examined by X-ray diffraction, as shown in Fig. 7. The generation of magnetite by
3.4. Effect of milling time and magnetic intensity on the magnetic separation of iron and manganese The influence of milling time and magnetic intensity on the magnetic separation of iron and manganese was studied on the basis of the above test work. It can be seen from Fig. 8 (A) and (B) that the recovery and grade of iron reached a plateau after milling for 20 min; however, the recovery and grade of manganese decreased slightly after milling for 30 min because grinding could have altered the crystal structure of manganese oxide in the roasted product to yield an amorphous mixture (Bid and Pradhan, 2003). In Fig. 8(C) and (D), both iron and manganese increased with increasing magnetic intensity, while the grade of both changed conversely. To yield an acceptable iron and manganese separation, the optimal magnetic intensity for iron separation was 1000–1200 G and that for manganese 11,000–12,000 G. The XRD patterns for both iron and manganese concentrates in Fig. 8(E) and (F) show that the main phases in the iron concentrate were Fe3O4 and SiO2 180
Minerals Engineering 126 (2018) 177–183
N. Peng et al.
°C
°C
Fig. 6. Effect of temperature (A, B), roasting duration (C, D) and CO concentration (E, F) on iron and manganese magnetic separation.
and those of the manganese concentrate were SiO2, MnO and Mn3O4. The roasted product was subjected to multi-step magnetic separation under the optimal magnetic intensity, and the element balance of iron and manganese across the whole process is shown in Table 2. The grade of iron and manganese reached 62.21% and 35.21%, respectively. 4. Conclusions This study showed that the decomposition of manganese ferrite MnFe2O4 is important for the recycling of high-iron-bearing manganese residues. An oxidative atmosphere favours the decomposition of manganese ferrite, which was confirmed by thermodynamic calculation and experiments. Haematite was further reduced to magnetite in a reducing roast, which caused the generation of manganous oxide and manganous-manganic oxide. In this process, the regeneration of manganese ferrite through the reaction between manganous oxide and haematite was found to be negligible. Iron and manganese in the roasted product were dissociated by a milling pretreatment and then recovered in a two-
Fig. 7. XRD spectrums of the reductively roasted products at various temperature. 181
Minerals Engineering 126 (2018) 177–183
N. Peng et al.
Fig. 8. Effect of milling time (A, B) and magnetic intensity (C, D) on the magnetic separation of iron (A, C) and manganese (B, D) and XRD patterns of the iron (E) and manganese (F) concentrates.
Acknowledgements
Table 2 Iron and manganese balance of whole process (wt%). Samples
Grade
Yield
Fe
Mn
16.79 62.21 8.45 2.51
15.12 4.26 35.21 1.87
The authors would like to thank the Natural Science Foundation of China (51574295) for financial support for this study.
Recovery Fe
Mn
100 72.29 19.61 6.21
100 5.50 90.75 5.14
References EMR Iron concentration Manganese concentration Tailings
100 19.51 38.97 41.52
Bid, S., Pradhan, S.K., 2003. Preparation of zinc ferrite by high-energy ball-milling and microstructure characterization by Rietveld’s analysis. Mater. Chem. Phys. 82 (1), 27–37. Das, A.P., Swain, S., Panda, S., Pradhan, N., Sukla, L.B., 2012. Reductive acid leaching of low grade manganese ores. Geomaterials 2 (04), 70. Du, B., Hou, D., Duan, N., Zhou, C., Wang, J., Dan, Z., 2015. Immobilization of high concentrations of soluble Mn(II) from electrolytic manganese solid waste using inorganic chemicals. Environ. Sci. Pollut. Res. 22 (10), 7782–7793. Duan, N., Fan, W., Changbo, Z., Chunlei, Z., Hongbing, Y., 2010. Analysis of pollution materials generated from electrolytic manganese industries in China. Resour. Conserv. Recycl. 54 (8), 506–511. Duan, N., Zhou, C., Chen, B., Jiang, W., Xin, B., 2011. Bioleaching of Mn from manganese residues by the mixed culture of Acidithiobacillus and mechanism. J. Chem. Technol.
step magnetic separation. The enrichment of iron and manganese and the production of a low-valence manganese oxide make this process promising and attractive for the recycling of high-iron-bearing manganese residues.
182
Minerals Engineering 126 (2018) 177–183
N. Peng et al.
Ag and In from high iron-bearing zinc calcine. J. Central South Univ. 24 (5), 1082–1089. Sorensen, B., Gaal, S., Ringdalen, E., Tangstad, M., Kononov, R., Ostrovski, O., 2010. Phase compositions of manganese ores and their change in the process of calcination. Int. J. Miner. Process. 94 (3), 101–110. Xin, B., Chen, B., Duan, N., Zhou, C., 2011. Extraction of manganese from electrolytic manganese residue by bioleaching. Bioresour. Technol. 102 (2), 1683–1687. Yan, S., Qiu, Y.-R., 2014. Preparation of electronic grade manganese sulfate from leaching solution of ferromanganese slag. Trans. Nonferrous Metals Soc. Chin. 24 (11), 3716–3721. Yao, J., Cheng, S., Xiao, S.-H., 2003. Influence of lixiviate rate of manganese in rhodochrosite by adding addition agent. J.-Jishou Univ. Nat. Sci. 24 (1), 43–45. Yuzhu, O., Xiaowei, P., Jianbing, C., Zhiping, L., Xiaodong, D., 2007. Ultrasonic leaching of electrolytic manganese residue with additive. Environ. Protect. Chem. Ind. 27 (3), 257–259. Zhou, C., Du, B., Wang, N., Chen, Z., 2014. Preparation and strength property of autoclaved bricks from electrolytic manganese residue. J. Cleaner Prod. 84, 707–714.
Biotechnol. 86 (6), 832–837. Elsherief, A., 2000. A study of the electroleaching of manganese ore. Hydrometallurgy 55 (3), 311–326. Källén, M., Hallberg, P., Rydén, M., Mattisson, T., Lyngfelt, A., 2014. Combined oxides of iron, manganese and silica as oxygen carriers for chemical-looping combustion. Fuel Process. Technol. 124, 87–96. Li, C., Zhong, H., Wang, S., Xue, J., 2014. Leaching behavior and risk assessment of heavy metals in a landfill of electrolytic manganese residue in Western Hunan, China. Hum. Ecol. Risk Assess.: Int. J. 20 (5), 1249–1263. Li, H., Zhang, Z., Tang, S., Li, Y., Zhang, Y., 2008. Ultrasonically assisted acid extraction of manganese from slag. Ultrason. Sonochem. 15 (4), 339–343. Liu, Y., Lin, Q., Li, L., Fu, J., Zhu, Z., Wang, C., Qian, D., 2014. Study on hydrometallurgical process and kinetics of manganese extraction from low-grade manganese carbonate ores. Int. J. Min. Sci. Technol. 24 (4), 567–571. Oberteuffer, J., 1974. Magnetic separation: a review of principles, devices, and applications. IEEE Trans. Magn. 10 (2), 223–238. Peng, B., Peng, N., Liu, H., Xue, K., Lin, D.-H., 2017. Comprehensive recovery of Fe, Zn,
183