Extraction of Ferronickel Concentrate by Reduction Roasting-Magnetic Separation from Low Grade Laterite Nickel Ore under the Action of Compound Additives

Abstract

Nickel is an important strategic metal in the world. As the sulfide ore containing high-grade nickel is increasingly depleted, the laterite nickel ore, which is rich in resources, has attracted people’s attention. In this paper, the reduction roasting-magnetic separation process is used to study the method of preparing ferronickel concentrates from low-grade laterite nickel ore under the action of composite additives (Na₂CO₃ and CaF₂). The research results showed that when the ratio of additives Na₂CO₃ and CaF₂ was 1:7, reduction temperature was 1250°C, reduction time was 60 min, magnetic field strength was 150 mT, and wet grinding time was 12 min, the nickel grade and recovery extent were 8.39 wt.% and 98.54%, iron grade and recovery extent were 67.70 wt.% and 71.73%.

Fig. 4 Experimental process.

1. Introduction

Nickel is one of the important strategic metal elements. It has high boiling point, high melting point, good chemical stability, corrosion resistance and oxidation resistance. It has a wide range of applications in the fields of national defense and military industry, chemical industry, metallurgy, aerospace, transportation and new materials.^1,2) According to the statistics of all nickel reserves, 30% exists as sulfide ores with the balance comprised of oxide ores. In nickel production, about 60% of the metal nickel was produced from sulfide ores, and laterite nickel ore only accounts for 40%.^3–5) The mineralogy of laterite nickel ore is complex and the grade of nickel and iron is low, so the treatment cost of conventional methods is high.⁶⁾ It is usually processed by high-pressure acid leaching, but a large amount of sulfuric acid is consumed under high temperature and high pressure.⁷⁾ As the amount of high grade nickel sulfide ore decreases, recovering nickel from laterite is the only way to maintain nickel production.⁸⁾ Therefore, there is an urgent need to develop an economical method for recovering nickel from laterite nickel ore.

The mineral phase composition of laterite nickel ore is complex, and nickel usually exists in the form of ultrafine inclusions in many mineral phases.⁹⁾ Therefore, effective purification through physical separation process is challenging.^10,11) The processing methods of low-grade laterite nickel ore include pyrometallurgy and hydrometallurgy. The representative process in pyrometallurgical technology is RKEF (Rotary Kiln-Electric Furnace) process, and its product is ferronickel concentrate, which is suitable for processing saprolite laterite ore.¹²⁾ The primary disadvantage of these processes is that they require substantial energy inputs, the energy consumption is 2 to 3 times higher than that of sulfide ore treatment. The combined treatment of low grade laterite nickel ore by fire and wet process is a research focus at present, mainly including chlorination separation magnetic separation, reduced sulfide roasting magnetic separation, direct reduction and magnetic separation.^13,14) Wang et al.¹⁵⁾ adopted direct reduction and magnetic separation technology to achieve the recovery extent of 75.70% nickel and 77.97% iron at the reduction temperature of 1500°C and the reduction time of 90 min. Tang et al.¹⁶⁾ used hydrogen and carbon monoxide as reducing agent and carried out reduction tests at 700∼1000°C. After magnetic separation, the grade of nickel and iron were 5.43 wt.% and 56.86 wt.%, respectively, and the recoveries extent were 84.38% and 53.76%, respectively. Yuan et al.¹⁷⁾ obtained deep reduction products under the conditions of reduction temperature 1275°C, reduction time 50 min and carbon coefficient 2.5, and obtained nickel–iron concentrate with nickel grade of 6.96 wt.% and iron grade of 34.74 wt.% by magnetic separation under the magnetic field strength of 72 kA/m, indicating that temperature affects direct reduction and magnetic separation. Although this method has been achieved in the laboratory, in industrial production, due to the limitations of equipment and production scale, industrialization has not yet been achieved. In order to further improve the enrichment effect of nickel and optimize the reduction and roasting conditions of laterite nickel ore, many scholars have studied the effects of different types of additives on promoting the reduction and roasting of laterite nickel ore to enrich nickel. Additives such as sulfur, chloride, sodium carbonate and fluoride were added in the roasting process and applied to reduction roasting and magnetic separation, and then the corresponding separation process was carried out. For example, Valix et al.¹⁸⁾ used carbon monoxide as the reducing agent and added the additive elemental sulfur to conduct the high temperature reduction experiment on limonite-type laterite nickel ore. In the presence of 5 wt.% S, the limonite-type laterite nickel ore was reduced at 600°C, and the nickel recovery extent reached 98%. Sulfur in the mixture promoted the aggregation and growth of ferronickel particles by reducing the surface tension of ferronickel particles. Sun et al.¹⁹⁾ studied the effects of S, Na₂S, Na₂SO₄ and the mixture of Na₂O and S on the grade and recovery of nickel. Adding 10 wt.% sulfur and 2 wt.% coal for roasting at 1200°C for 50 min, the grade of nickel in ferronickel concentrate obtained by magnetic separation was 7.21 wt.%, while the recovery extent of iron sharply decreased from 62.39% to 29.43%. Because the change of nickel grade had a similar trend to the FeS diffraction peak in the presence of Na₂S, Na₂SO₄ and the mixture of Na₂O and S, it was proposed that Na₂SO₄ was decomposed into Na₂S or S and reacts with FeO and Fe to form FeS, respectively. In addition, Na₃MgAl (SiO₄)₂ produced by Na₂O reacts with silicates to promote the migration of metal grains. Zhou et al.²⁰⁾ studied the mechanism of nickel enrichment by sodium chloride enhanced reduction roasting of laterite nickel ore, and the results showed that the addition of sodium chloride was helpful to the aggregation and growth of ferronickel particles. When 10 wt.% sodium chloride was added as additive to reduce the low-grade laterite nickel ore, a ferronickel concentrate with 7.09 wt.% nickel and a nickel recovery extent of 98.31% was obtained. Guo et al.²¹⁾ obtained high recovery extent of nickel and cobalt by basic roasting-water or acid leaching process, and could remove impurities such as chromium, aluminum and silicon at low temperature; the thermodynamic data of Na₂CO₃ enhanced reduction roasting of brown iron laterite showed that Na₂CO₃ could react with Cr₂O₃, Al₂O₃ and SiO₂ within the experimental temperature range. In summary, the pyrometallurgical treatment of laterite nickel ore to produce iron nickel alloy particles has the characteristics of short process flow and high efficiency.^22,23)

In this paper, the low-grade laterite nickel ore was selected as the research object, and the process of reduction roasting-magnetic separation of ferronickel concentrate was adopted, and focused on the influence of compound additives on the reduction roasting process of laterite nickel ore in order to efficiently enrich low-grade laterite nickel ore, the process was short, low energy consumption, good economic and environmental benefits.

2. Material and Methods

2.1 Experimental materials

The raw material used in the experiment was silicomagnesite laterite nickel ore. The chemical analysis results are shown in Table 1. The laterite nickel ore contained 1.06 wt.% nickel and 11.56 wt.% iron. The main impurities were SiO₂ and MgO, the contents of which were 29.80% and 35.41% respectively. It was a typical ore with high content silicon, magnesium and low-grade nickel, suitable for smelting by pyrometallurgy process.

Table 1 Chemical composition of laterite nickel ore.

The results of X-ray diffraction (XRD) analysis of the raw ore are shown in Fig. 1. According to the XRD analysis, the laterite nickel ore was mainly composed of silicate and oxide minerals containing hydroxyl. Iron mainly existed in minerals in the form of iron oxide, and magnesium mainly existed in serpentine in the form of magnesium oxide. Due to the low nickel content in the laterite nickel ore, the phase of nickel was not detected, but the properties of nickel and iron were similar. Combined with scanning electron microscope (SEM) Fig. 2, it could be seen that nickel coincided with silicon, magnesium and oxygen, and it was concluded that nickel mainly existed in serpentine in the form of nickel oxide.

Fig. 1

XRD analysis of laterite nickel ore.

Fig. 2

Electron microscopic imaging of laterite nickel ore.

Through scanning electron microscope and energy spectrum analysis, the micro morphology and element content of the raw ore are obtained, and the distribution of each element in the raw ore is further studied, as shown in Fig. 2 and Fig. 3. The observation showed that the distribution areas of oxygen, silicon and magnesium in the raw ore were highly coincident, and the aggregated particles were large, accounting for most of the area; According to the EDS diagram, the main elements in the laterite nickel ore were iron, magnesium and silicon, as well as a small amount of nickel and aluminum; According to the results of XRD (Fig. 1), the main phase of the laterite nickel ore was serpentine. The coincidence degree between the distribution area of iron element and oxygen element was also high. It could be concluded that some iron in laterite nickel ore existed in the form of iron oxide. Then, according to the distribution area of iron element and the depth of color, the content of iron oxide was low, which corresponded to the analysis results of chemical elements of raw ore. The distribution of nickel element was dispersible, evenly distributed in the whole mineral, and the imaging color was light. The content of nickel element was low, and the distribution coincidence degree with oxygen, silicon, magnesium is high, and there was a small area of overlap with iron. According to the analysis, a small part of nickel replaces the position of iron in iron oxide in the form of isomorphism, and coexists with iron, most nickel exists in serpentine.

Fig. 3

Energy spectrum analysis of laterite nickel ore.

2.2 Reducing agent and additive

The reducing agent used in this experiment was low-cost anthracite. After crushing and ball milling, more than 80% of the particle size was less than 74 µm. The chemical analysis results are shown in Table 2. The additive used in the experiment is the compound additive of calcium fluoride and sodium carbonate. The component analysis is shown in Table 3 and Table 4.

Table 2 Industrial analysis of anthracite (wt.%, ω).

Table 3 Composition analysis of CaF₂ (wt.%, ω).

Table 4 Composition analysis of Na₂CO₃ (wt.%, ω).

2.3 Method

This paper studied the reduction reaction of low-grade laterite nickel ore under the condition of adding composite additives (Na₂CO₃ and CaF₂) and reducing agent (anthracite) in high temperature environment and protective gas (nitrogen) atmosphere. The purpose was to use additives to destroy the serpentine structure and changed the valuable metals such as nickel and iron in minerals into simple oxides. The addition of reducing agent in minerals could reduce the nickel and iron oxides in laterite nickel ore into magnetic nickel and iron metals. Finally, the reduced nickel and iron metals were separated from gangue by magnetic separation. In this experiment, the effects of additive dosage and ratio (Na₂CO₃:CaF₂), reduction roasting temperature, reduction roasting time, wet grinding time and magnetic separation intensity on the reduction of nickel and iron in laterite nickel ore were studied, and finally the best experimental conditions were determined to realize the efficient enrichment of nickel and iron in laterite nickel ore. The experimental process is shown in Fig. 4.

Fig. 4

Experimental process.

3. Results and Discussion

3.1 The effect of the amount of compound additives

The effect of different additive ratios (Na₂CO₃:CaF₂) on the reduction roasting of laterite nickel ore and magnetic separation of ferronickel is shown in Fig. 5.

Fig. 5

The influence of different additive ratios on nickel and iron grades and recovery extent.

As can be seen from Fig. 5, the grade and recovery extent of nickel firstly increased and then decreased with the increase of additive ratio (Na₂CO₃:CaF₂), when the addition amount of Na₂CO₃ was 0 wt.% and the addition amount of CaF₂ was 8 wt.%, the grade of nickel was 4.68 wt.% and the grade of iron was 45.04 wt.%. When the addition amount of Na₂CO₃ was 1 wt.% and the addition amount of CaF₂ was 7 wt.%, the grade and recovery extent of nickel were 8.39 wt.% and 98.54%, respectively. At this time, the grade and recovery extent of iron were 67.70 wt.% and 71.73%, respectively. When the ratio of additives (Na₂CO₃:CaF₂) continued to increase, that was, the content of Na₂CO₃ in the additive increased and the content of CaF₂ decreased. Excessive Na₂CO₃ would inhibit the progress of the reduction reaction, the addition of CaF₂ would promote the aggregation of nickel and iron, so that the grades of nickel and iron decreased instead. Considering comprehensively, the best additive ratio (Na₂CO₃:CaF₂) is 1:7.

3.2 The effect of reduction time

The effect of reduction time on laterite nickel ore reduction roasting and magnetic separation of ferronickel is shown in Fig. 6.

Fig. 6

The effect of reduction time on nickel and iron grades and recovery extent.

As shown in Fig. 6, with the extension of reduction time, the grade of nickel increased from 7.16 wt.% to 8.39 wt.% and then decreased to 6.44 wt.%, the recovery extent of nickel increased from 85.06% to 98.54% and then decreased to 79.96%. When the reduction time was 20 min, the grade of nickel and iron were 7.16 wt.% and 85.06 wt.%, respectively, and the recovery extent were 51.88% and 64.47%, respectively. Due to the short reduction time, the aggregation and growth time of nickel and iron particles were insufficient, and the ferronickel particles were small and magnetism was very weak, which couldn’t be enriched effectively in the magnetic separation process. Therefore, the extension of the reduction time is beneficial to the magnetic separation. When the reduction time was 60 min, the grade and recovery extent of nickel were the highest, which were 8.39 wt.% and 98.54%, respectively, and the grade and recovery extent of iron were 67.70 wt.% and 71.73%, respectively. However, when the reduction time was 80 min and 100 min, the grade and recovery extent of nickel and iron had a downward trend. The excessive metal iron diluted the nickel in the ferronickel alloy, resulted in a decrease in the nickel grade. Considering factors such as ferronickel grade, recovery extent and energy consumption, the best roasting time is 60 min.

3.3 The effect of reduction temperature

The effect of reduction temperature on laterite nickel ore reduction roasting and magnetic separation of ferronickel is shown in Fig. 7.

Fig. 7

The effect of reduction temperature on nickel and iron grades and recovery extent.

As can be seen from Fig. 7, with the increase of reduction temperature, the grade and recovery extent of iron gradually increased, while the grade and recovery extent of nickel firstly increased and then decreased. When the reduction roasting temperature increased to 1250°C, the recovery extent of nickel began to show a downward trend, and some ferric olivine was produced in the raw ore at this temperature, which affected the polymerization and growth of ferronickel alloy. When the reduction roasting temperature continued to rise to 1300°C, the recovery extent of nickel decreased from 98.54% to 89.24%. Too high reduction temperature would lead to mineral sintering, inhibited the reduction process, hindered the formation of nickel-ferric metal phase during the reaction process, therefore, the nickel grade in ferronickel concentrate would be reduced. At 1250°C, the grade and recovery extent of nickel reached the highest values of 8.39 wt.% and 98.54%, respectively, and the grade and recovery extent of iron were 67.70 wt.% and 71.73%, respectively. With the increase of the temperature of reduction roasting, the activity of reactants is enhanced, the reaction extent is also accelerated, and nickel and iron oxides in minerals are reduced to nickel and iron metals. Therefore, reduction roasting temperature is an important factor which could affect the enrichment of ferronickel in laterite. Considering the grade and recovery extent of nickel and iron, 1250°C was chosen as the optimal reduction temperature.

3.4 The effect of magnetic field intensity

The effect of magnetic field intensity on laterite nickel ore reduction roasting-magnetic separation of ferronickel is shown in Fig. 8.

Fig. 8

Effect of magnetic field intensity on the grade and recovery extent of nickel and iron.

Because of the magnetic properties of nickel and iron, the magnetic field intensity can determine the grade and recovery extent of ferronickel. Ferronickel can be separated from impurity ions by magnetic separator to obtain ferronickel concentrate. When the magnetic field intensity is too low, the recovery extent of ferronickel will be reduced, and some ferronickel will enter the magnetic separation tailings; When the magnetic field is too high, the impurities will be adsorbed together, which will reduce the grade of ferronickel in the concentrate. With the increase of magnetic field intensity, the grade of nickel increased from 6.64 wt.% to 8.39 wt.% and then decreased to 5.80 wt.%, and the recovery extent increased from 81.29% to 98.54% and then decreased to 84.81%. When the magnetic field intensity was 150 mT, the grade and recovery extent of nickel reached the highest value of 8.39 wt.% and 98.54% respectively, and the grade and recovery extent of iron were 67.70 wt.% and 71.73% respectively. Considering the grade and recovery of nickel, 150 mT is the best magnetic separation intensity.

3.5 The effect of wet grinding time

The effect of wet grinding time on reduction roasting magnetic separation of ferronickel from laterite nickel ore is shown in Fig. 9.

Fig. 9

Effect of wet grinding time on the grade and recovery extent of nickel and iron.

As can be seen from Fig. 9, wet grinding time has a certain influence on the grade and recovery of nickel and iron. When the wet grinding time was 3 min, the grade and recovery extent of nickel and iron are both low. When the wet grinding time reached 12 min, the grade and recovery extent of nickel also reached the highest, which were 8.39 wt.% and 98.54% respectively; continued to extend the wet milling time, and the calcined product would be ground into extremely fine powder. Due to the smaller powder size, the corresponding specific surface area and surface energy would increase. The thermodynamic properties of this system are unstable, and the entire fine powder system will aggregate and reduce small free enthalpy to ensure stability, the fine powder will form particles again, which is not conducive to magnetic separation. Therefore, due to the long wet grinding time, the ferronickel particles that are well polymerized during the reduction and roasting process will be destroyed. When the wet grinding time was 12 minutes, the nickel grade and recovery extent reached the highest, which were 8.39 wt.% and 98.54%, respectively. At this time, the iron grade and recovery extent were 67.70 wt.% and 71.73%, respectively. Grinding fineness determines the monomer dissociation between nickel–iron minerals and gangue minerals and between nickel and iron particles is sufficient or not. Considering comprehensively, the optimal wet grinding time is selected as 12 minutes.

3.6 Product analysis

XRD analysis and chemical analysis were performed on the product ferronickel concentrate, and the results are shown in Fig. 10 and Table 5. As can be seen from Table 5 that the nickel content is 8.39 wt.%, and the iron content is 67.70 wt.%; it can be seen from Fig. 10 that it is mainly a ferronickel phase.

Fig. 10

XRD analysis of ferronickel concentrate.

Table 5 XRD chemical analysis of nickel–iron concentrate.

The grade of nickel in the magnetic separation slag is 0.03%, and the grade of iron is 2.85%. The reasons for containing nickel and iron in the magnetic separation slag are as follows: first, the nickel and iron are not completely reduced in the reduction stage; second, the ferronickel particles after fine grinding are too small and trapped in the slag, which are carried away by water flow and fail to be enriched in the magnetic separation stage. The theoretical maximum grade of nickel and iron in the concentrate is 8.49% and 70.80%, respectively. The formula is as follows:   

\begin{equation*} \omega = \frac{m1 \times \omega 1 + m2 \times \omega 2}{m1 + m2} \end{equation*}

In the formula, ω is the theoretical maximum value of nickel/iron, M1 is the concentrate quality, ω1 is the concentrate nickel/iron grade, M2 is the magnetic separation slag quality, ω2 is the magnetic separation slag nickel/iron grade.

3.7 Phase transformation during the roasting of laterite nickel ore

In order to explain the effect of additives on the reduction roasting-magnetic separation of laterite nickel ore to prepare ferronickel concentrate from the microscopic level, and to reveal the regular of promoting the aggregation and growth of ferronickel particles to form ferronickel concentrate, scanning electron microscopy was performed on the roasted products under different additive conditions (SEM) and X-ray diffraction analysis (XRD), as shown in Fig. 11 and Fig. 12. The element analysis of each point in the figure is shown in Table 6.

Fig. 11

SEM image of roasted sample with accelerator added.

Fig. 12

SEM image and EDS analyses of without additives.

Table 6 Elemental analysis of laterite nickel ore roasted with Na₂CO₃ and CaF₂.

Figure 11 and Fig. 12(a) are the images of scanning electron microscope (SEM) at 1250°C with and without additives. It can be observed that when additives are added, there are obviously large ferronickel particles, and when there is no additive, only small ferronickel particles can be observed. Figures 12(b) and (c) show the energy spectrum analysis of points 1 and 2 in (a). The main component of point 1 is iron, and point 2 is basically the same as the raw ore, indicating that some areas of the sample without additives have not been reduced, and the reduced ferronickel agglomeration effect is not obvious. This indicates that when Na₂CO₃ and CaF₂ are added, the serpentine structure can be effectively destroyed and the nickel and iron components in the raw ore are released to form ferronickel.

As can be seen from Fig. 13 and Fig. 14 that iron mainly exists in Mafic olivine when no additives are added. When Na₂CO₃ and CaF₂^24,25) are added, CaF₂ is conducive to the dissociation of iron and nickel silicate into simple oxides, improve the activity of metal oxides and improve their reduction progress; Na₂CO₃ can be decomposed into Na₂O and CO₂ at 851°C, in which Na₂O and CaF₂ react with Fe₂SiO₄ to form (Na, Ca)₂(Fe, Mg)₅Si₈O₂₂(OH)₂ phase with low melting point, which changes the structure of silicate minerals from island to chain and improves the reaction activity of silicate minerals. And Na₂O also reacts with Cr₂O₃ and Al₂O₃ to form sodium chromate and sodium aluminate, which can effectively remove impurity ions such as Cr, Si and Al. Therefore, when Na₂CO₃ and CaF₂ are added, the activity of the metal oxide can be increased, the reduction progress can be improved, the aggregation and growth of nickel–iron crystal grains can be promoted, and the recovery extent of Ni can be improved.

Fig. 13

XRD pattern of the reduction and roasting product of laterite nickel ore under the action of compound additives.

Fig. 14

XRD pattern of laterite nickel ore reduction roasting without additives.

4. Conclusion

1. (1)    The low-grade laterite nickel ore contained 1.06 wt.% nickel and 11.56 wt.% iron. It was mainly composed of hydroxyl-containing silicate and oxide minerals. Among them, iron mainly existed in minerals in the form of iron oxides and magnesium mainly existed in the serpentine in the form of magnesium oxide, the nickel content was too low to be detected in XRD.
2. (2)    Silicomagnesite laterite nickel ore was directly reduced and roasted without additives. The reduction effect of nickel and iron was not obvious. The effect could be improved obviously by adding CaF₂ and Na₂CO₃. Under the conditions of total compound addition of 8 wt.%, compound additive ratio (Na₂CO₃:CaF₂) of 1:7, reduction time of 60 min, reduction temperature of 1250°C, magnetic separation strength of 150 mT and wet grinding time of 12 min, the grades of nickel and iron in ferronickel concentrate were 8.39 wt.% and 67.70 wt.% respectively, and the recovery extent of nickel and iron were 98.54% and 71.73% respectively.

Acknowledgements

Financial support for this study was supplied from the Yunnan Provincial Key Research and Development Program - International Science and Technology Cooperation Special Project (Project Nos. 2018IA055) and the National Natural Science Foundation of China (Project Nos. 52074140).

REFERENCES

• 1)   B.  Li,  Z.G.  Ding,  Y.G.  Wei,  S.W.  Zhou and  H.  Wang: Mater. Trans. 59 (2018) 1180–1185. doi:10.2320/matertrans.M2017351
• 2)   S.  Pintowantoro and  F.  Abdul: Mater. Trans. 60 (2019) 2245–2254. doi:10.2320/matertrans.MT-M2019101
• 3)   J.  Yang,  G.Q.  Zhang,  O.  Ostrovski and  S.  Jahanshahi: Miner. Eng. 54 (2013) 110–115. doi:10.1016/j.mineng.2013.05.009
• 4)   G.M.  Mudd: Ore Geol. Rev. 38 (2010) 9–26. doi:10.1016/j.oregeorev.2010.05.003
• 5)   A.  Nosrati,  K.  Quast,  D.F.  Xu,  W.  Skinner,  D.J.  Robinson and  J.  Addai-Mensah: Hydrometallurgy 146 (2014) 29–39. doi:10.1016/j.hydromet.2014.03.004
• 6)   D.F.  Xu,  L.X.  Liu,  K.  Quast,  J.  Addai-Mensah and  D.J.  Robinson: Adv. Powder Technol. 24 (2013) 750–756. doi:10.1016/j.apt.2013.03.013
• 7)   A.  Jordens,  Y.P.  Cheng and  K.E.  Waters: Miner. Eng. 41 (2013) 97–114. doi:10.1016/j.mineng.2012.10.017
• 8)   D.Q.  Zhu,  Y.  Cui,  S.  Hapugoda,  K.  Vining and  J.  Pan: Trans. Nonferrous Met. Soc. China 22 (2012) 907–916. doi:10.1016/S1003-6326(11)61264-8
• 9)   H.R.  Watling,  A.D.  Elliot,  H.M.  Fletcher,  D.J.  Robinson and  D.M.  Sully: Aust. J. Earth Sci. 58 (2011) 725–744. doi:10.1080/08120099.2011.602986
• 10)   S.  Farrokhpay,  M.  Cathelineau,  S.B.  Blancher,  O.  Laugier and  L.  Filippov: J. Geochem. Explor. 196 (2019) 270–281. doi:10.1016/j.gexplo.2018.11.002
• 11)   K.  Quast,  J.N.  Connor,  W.  Skinner,  D.J.  Robinson and  J.  Addai-Mensah: Miner. Eng. 79 (2015) 261–268. doi:10.1016/j.mineng.2015.03.017
• 12)   J.  Lu,  S.J.  Liu,  J.  Shangguan,  W.G.  Du,  F.  Pan and  S.  Yang: Miner. Eng. 49 (2013) 154–164. doi:10.1016/j.mineng.2013.05.023
• 13)   B.  Li,  H.  Wang and  Y.G.  Wei: Miner. Eng. 24 (2011) 1556–1562. doi:10.1016/j.mineng.2011.08.006
• 14)   S.  Farrokhpay and  L.  Filippov: Int. J. Miner. Process. 151 (2016) 59–67. doi:10.1016/j.minpro.2016.04.007
• 15)   L.W.  Wang,  X.M.  Lü,  M.  Liu,  Z.X.  You,  X.W.  Lüand and  C.G.  Bai: Int. J. Miner. Metall. Mater. 25 (2018) 744–751. doi:10.1007/s12613-018-1622-7
• 16)   X.H.  Tang,  R.Z.  Liu,  L.  Yao,  Z.J.  Ji,  Y.T.  Zhang and  S.Q.  Li: Int. J. Miner. Metall. Mater. 21 (2014) 955–961. doi:10.1007/s12613-014-0995-5
• 17)   S.  Yuan,  W.T.  Zhou,  Y.J.  Li and  Y.X.  Han: Trans. Nonferrous Met. Soc. China 30 (2020) 812–822. doi:10.1016/S1003-6326(20)65256-6
• 18)   M.  Valix and  W.H.  Cheung: Miner. Eng. 15 (2002) 607–612. doi:10.1016/S0892-6875(02)00068-7
• 19)   T.C.  Sun,  M.  Jiang,  Z.G.  Liu,  N.  Liu,  S.Y.  Zhang,  J.  Kou and  C.Y.  Xu: J. China Univ. Min. Technol. 42 (2013) 838–844. doi:10.13247/j.cnki.jcumt.2013.05.021
• 20)   S.W.  Zhou,  Y.G.  Wei,  B.  Li,  H.  Wang,  B.Z.  Ma and  C.Y.  Wang: Metall. Mater. Trans. B 47 (2016) 145–153. doi:10.1007/s11663-015-0478-8
• 21)   Q.  Guo,  J.K.  Qu,  T.  Qi,  G.Y.  Wei and  B.B.  Han: Miner. Eng. 24 (2011) 825–832. doi:10.1016/j.mineng.2011.03.001
• 22)   M.J.  Rao,  G.H.  Li,  X.  Zhang,  J.  Luo,  Z.W.  Peng and  T.  Jiang: Sep. Sci. Technol. 51 (2016) 1408–1420. doi:10.1080/01496395.2016.1162173
• 23)   X.P.  Wang,  T.C.  Sun,  C.  Chen and  J.  Kou: Int. J. Miner. Metall. Mater. 25 (2018) 383–390. doi:10.1007/s12613-018-1582-y
• 24)   Q.  Guo,  J.K.  Qu,  B.B.  Han,  G.Y.  Wei,  P.Y.  Zhang and  T.  Qi: Trans. Nonferrous Met. Soc. China 24 (2014) 3979–3986. doi:10.1016/S1003-6326(14)63559-7
• 25)   Z.L.  Xue,  C.P.  Xiao,  G.H.  Hang,  T.Y.  Xiong and  Y.C.  Xie: J. Chongqing Univ. 40(12) (2017) 43–51. doi:10.11835/j.issn.1000-582X.2017.12.006