Advertisement

Mining, Metallurgy & Exploration

, Volume 36, Issue 1, pp 131–138 | Cite as

On the Mechanisms of Silica (SiO2) Recovery in Magnetite Ore Low-Magnetic-Drum Concentration

  • M. Llamas-Bueno
  • A. López-ValdiviesoEmail author
  • M. A. Corona-Arroyo
Article
  • 23 Downloads

Abstract

Magnetite (Fe3O4) ore is concentrated by low-magnetic-field drums to recover a magnetite concentrate that is low in silica (SiO2). The presence of SiO2 in the magnetite concentrate for steel production increases the steel processing costs, so a major challenge in magnetic concentration is to lower the SiO2 grade in the Fe concentrate. This work presents studies that were carried out on the removal of SiO2 from the magnetite concentration at the plant scale. Studies were performed with a three 36 × 96″ drum unit processing 46 ton/h of rougher magnetite concentrate. These studies showed that silica appears in the magnetite concentrate by three mechanisms, namely SiO2 entrainment in Fe3O4 chains, heterocoagulation between Fe3O4 and SiO2 particles, and mineral locking of SiO2 to Fe3O4. The SiO2 entrainment mechanism had the highest contribution (75%) to the SiO2 recovered in the Fe concentrate. Electrokinetic studies on Fe3O4 and SiO2 showed that heterocoagulation occurs because of the low negative zeta potential of the minerals at the pH of the plant slurry. This study also showed that a high percentage of ultrafine size SiO2 particles trapped in Fe3O4 agglomerates was not removed by the dilution water for the slurry fed to the drum. To lower the SiO2 recovery in the drum magnetic concentration process, work should be directed towards the removal of trapped SiO2 in Fe3O4 agglomerates.

Keywords

Magnetite Magnetic concentration Silica Iron ore Iron concentrate 

1 Introduction

Silica (SiO2) is a major impurity in iron concentrate in steelmaking and highly impacts the production costs of steelmaking. Silica has to be melted, so more energy is required to make the steel. In addition, silica is responsible for the addition of more additives, iron loss in slags, and high furnace refractory consumption during the steelmaking process. Accordingly, a challenge in iron ore processing is to produce iron concentrates with a low silica grade. SiO2 is present in Fe concentrates as quartz (SiO2) and SiO2-bare minerals.

Magnetite (Fe3O4) ore processing is carried out with low-intensity magnetic rotating drums [1, 2, 3]. This is most commonly used type of magnetic separator used in the world [2, 3]. Huge tonnages of low-grade magnetite ore are processed by this means. A finely ground-magnetite slurry is fed to the bottom of a tank with a drum with many magnetic bars inside it with magnetic fields ranging from 0.03 to 0.1500 T [4]. Plenty of water is added to further dilute the slurry. This water is used to decrease the entrainment of diamagnetic minerals (quartz, silica-bare minerals, carbonate, and sulfides) in magnetite agglomerates and to wash out the diamagnetic minerals from the magnetite agglomerates attached to the surface of the rotating drum. As soon as they enter the tank, the magnetite particles are subjected to a magnetic field and travel towards the rotating drum due to the magnetic field gradient between the drum surface and the tank bottom. The magnetite particles form agglomerates, which are attached on the drum surface. These magnetite agglomerates (Fe concentrate) are moved by the drum out of the magnetic field region and discharged from of the tank. On the other hand, the diamagnetic minerals (tailings) are dragged by water from the tank bottom. The iron concentrate passes through several rotating drums in order to maximize the removal of diamagnetic minerals and produce an iron concentrate with minimal contents of SiO2, carbonates, and sulfides. Even though the process of magnetic concentration by rotating drums is a very old technology, not much work has been undertaken on the mechanisms responsible for the appearance of silica-bare minerals and other diamagnetic minerals in the iron concentrate.

This work aimed to determine the mechanisms through which SiO2 appears in the Fe concentrate in the concentration process by low-intensity magnetic rotating drums. A detailed analysis was undertaken on the SiO2 recovery in a continuous plant unit with three rotating drums. Plant studies were complemented with laboratory studies on the behavior of magnetite particles in a uniform external magnetic field, the electrokinetics of quartz and magnetite, and the mineralogical characterization of the Fe concentrate.

2 Experimental

2.1 Magnetic Drum Unit

A magnetic drum unit in an iron magnetic concentration plant in Mexico was selected to assess the recovery of SiO2 in the Fe concentrate. This magnetic drum unit treated 46 t/h of a rougher Fe concentrate (63.1% Fe and 6% SiO2) with a P80 of 45 μm and had three 36 × 96″ drums rotating at 38.5 rpm. Each drum had 6 magnetic bars, each with a magnetic field of 0.12 T on the surface. The separation distance between the drum surface and the tank bottom was 2.5 in. In this unit, the Fe concentrate of each drum passed to the next one for further cleaning, so a final concentrate was obtained from the third drum. Dilution water was supplied to the slurry before it entered to the tank. A water flow meter was installed at the pipe of the dilution water to monitor the water flow supplied to the slurry. A schematic of the magnetic drum unit is presented in Fig. 1. The slurry pH was 6.5.
Fig. 1

Schematic of the magnetic drum unit

Samples of the feed, concentrate, and tailings of each of the three drums were taken at distinct dilution-water flow rates for a period of 4 h. For each dilution-water flow rate, a water and solid mass balance was carried out, and analyses of the Fe and SiO2 content, size distribution, and mineralogical properties were completed for the feed, concentrate, and tailings in order to determine the Fe and SiO2 overall recovery and their recoveries size-by-size fraction.

2.2 Zeta Potential Determination

Quartz crystals from Brazil and pure natural magnetite from Mexico were used for electrokinetic studies and to study the interactions between magnetite and quartz particles. First, the quartz and magnetite crystals were crushed and then finely ground using an agate mortar and pestle. Then, − 635 mesh size particles (− 20 μm) were collected for electrokinetic measurements, which were carried out using a Coulter Delsa 440 zetameter. For these measurements, a mineral/water suspension was prepared at 0.1%w solid concentration. The pH of the suspension was adjusted using dilute solutions of HCl and NaOH, and the suspension was conditioned at the desired pH for 30 min prior to placing it into the zetameter cell. The ionic strength of the suspension was kept constant at 0.01 mol/L for all the electrokinetic tests, and NaCl was used as the supporting electrolyte. The zeta potential of quartz and magnetite was determined in the absence and presence of sodium hexametaphosphate (Na6(PO3)6 as a function of pH. The minerals were contacted with the dispersant at the desired pH for 30 min prior to the measurement of their electrophoretic mobilities, which were transformed to zeta potentials using the Smolochowski equation.

3 Results and Discussion

3.1 Drum Magnetic Concentration

The SiO2 recovery and grade in the Fe concentrate of the second and third drums of the magnetic concentration unit were determined as functions of the dilution-water flow rate in the second drum while keeping the dilution-water flow rate in the third drum constant at 25 m3/h. Figures 2 and 3 show that the recovery and grade of SiO2 in the Fe concentrates decreased with the dilution-water flow rate up to a given point, above which the SiO2 grade and recovery did not lower further and a plateau was reached.
Fig. 2

SiO2 recovery in the Fe concentrate of the second and third drums as a function of the dilution-water flow rate in the second drum. The dilution-water flow rate in the third drum was constant at 25 m3/h

Fig. 3

SiO2 grade in the Fe concentrate of the second and third drums as a function of the dilution-water flow rate in the second drum. The dilution-water flow rate in the third drum was constant at 25 m3/h

Figure 4 shows the SiO2 yield in ton/h in the Fe concentrate of the second drum as a function of the particle size for two dilution-water flow rates, namely 24 and 60 m3/h. As noted, the yield of the SiO2 particles smaller than 20 μm decreased with the highest dilution-water flow rate. The SiO2 recovery in the Fe concentrate of the second drum as a function of particle size for the two dilution-water flow rates is presented in Fig. 5. This SiO2 recovery was determined from the SiO2 supplied to the second drum by the first drum. It can be seen that the recovery of SiO2 particles smaller than 20 μm are lower for the water flow of 60 m3/h. At this water flow rate, the hydrodynamic forces increased dragging more fine particles from the tank [4]. In addition, these forces likely removed some fine SiO2 particles trapped in Fe3O4 agglomerates. Accordingly, the recovery of the < 20 μm SiO2 particles decreased. For the 60 m3/h water flow rate, a high percentage of ultrafine SiO2 particles (< 10 μm) were still recovered in the Fe concentrate; however, as noted in Fig. 5, the recovery was 41% for the SiO2 particles smaller than 7 μm.
Fig. 4

SiO2 yield in the Fe concentrate of the second drum as a function of the particle size for two dilution-water flow rates in the second drum, namely 24 and 60 m3/h

Fig. 5

SiO2 recovery in the Fe concentrate of the second drum as a function of particle size for two dilution-water flow rates in the second drum, namely 24 and 60 m3/h

3.2 SiO2–Fe3O4 Locking in the Fe Concentrate

A size-by-size fraction SiO2 and Fe3O4 mass distribution and mineral liberation studies were carried out on the Fe concentrate going to the second drum. Each size fraction was mounted in an epoxy probe and the surface of the mineral particles was polished and coated with a gold thin film to make them conductive. The mineral liberation analysis was performed using a scanning electron microscope equipped with an energy dispersive X-ray spectrometer as reported elsewhere [5]. About 3000 particles were analyzed. Table 1 shows the SiO2 and Fe grade as well as the percentage of free SiO2 and Fe3O4 particles on each size fractions of the Fe concentrate.
Table 1

SiO2 and Fe grade and free SiO2 and Fe3O4 particles on each size fractions of the Fe concentrate supplied to the second drum

   

Grade

Free particles, %

 

Size

Wt

%

Relative*

Absolute+

Mesh

μm

%

Fe

SiO2

Fe3O4

SiO2

Fe3O4

SiO2

150

106

0.2

35.2

31.5

44.0

15.0

0.1

0.2

200

75

0.4

36.9

26.7

56.0

18.0

0.1

0.4

270

53

2.2

40.3

17.5

87.0

22.0

1.2

2.4

325

45

3.1

45.7

11.8

89.0

33.0

1.9

4.2

400

38

3.4

55.8

7.8

90.0

37.0

2.6

3.2

500

25

10.7

65.5

2.7

92.0

42.0

9.8

4.5

600

20

13.0

67.8

2.3

92.0

47.0

12.4

4.1

− 600

− 20

67.0

67.7

2.5

93.0

51.0

64.3

23.0

Total

 

100.0

65.6

3.4

  

92.4

41.9

*Relative free mineral is the percentage of free mineral particles in the population of mineral particles (free and locked) in the size fraction. +Absolute free mineral is the percentage of free mineral in the mass of the size fraction in the Fe concentrate. It was calculated from the relative free mineral times the mineral distribution in the size fraction

Table 1 shows that about 42% of the SiO2-bare mineral particles were free in the Fe concentrate that was supplied to the second drum. If all these free SiO2–bare mineral particles were removed in the magnetic concentration, the final Fe concentrate would have a grade of 1.6% SiO2. However, the Fe concentrates of the second and third drums had a grade of 2.9 and 2.7% SiO2, respectively. This indicates that SiO2–Fe3O4 locked and free SiO2–bare mineral particles were recovered in the Fe concentrate by the drums. Figure 6 shows the type of locking between the SiO2 and Fe3O4 particles in the Fe concentrate. The locked SiO2 were transported to the drum surface by the F3O4 associated to the SiO2. The amount of SiO2 due to SiO2–Fe3O4 locked particles was calculated to be 20% of the SiO2 in the Fe concentrate.
Fig. 6

SEM photomicrographs of free SiO2 and Fe3O4 particles and locked SiO2–Fe3O4 particles in Fe concentrate

3.3 SiO2–Fe3O4 Heterocoagulation

Scanning electron microscopy studies on the Fe concentrate showed that heterocoagulation occurred between the Fe3O4 and SiO2 particles during the magnetic concentration process. Figure 7 depicts SEM images of particles in the Fe concentrate, and it can be seen that SiO2 ultrafine particles attached to the surfaces of Fe3O4 coarse particles, and Fe3O4 ultrafine particles attached to the surfaces of coarse SiO2.
Fig. 7

SEM photomicrographs of Fe concentrate particles showing heterocoagulation between SiO2 and Fe3O4 particles

Heterocoagulation between particles is largely determined by the electric charge at the particle/aqueous solution interface, so the zeta potentials of magnetite and quartz were determined in order to delineate the interaction energy between the Fe3O4 and quartz (SiO2) particles. Quartz represents the SiO2-bare minerals. Figure 8 shows the zeta potentials of Fe3O4 and SiO2 as a function of the pH. The arrow in the figure indicates the plant slurry’s pH 6.5. At this pH value, the zeta potentials of Fe3O4 and SiO2 were − 12 and − 38.3 mV, respectively.
Fig. 8

Zeta potentials of SiO2 and Fe3O4 as functions of the pH at an ionic strength of 0.01 M. The arrow indicates the process slurry’s pH

The interaction energy (VT) between the Fe3O4 and SiO2 particles was determined using the Hogg–Healy–Fuerstenau model [6]. The magnetite particle was considered to be a plate and the SiO2 particle a sphere. VT takes into account the attractive van der Waals forces (VA) and the repulsive force (VR) due to the electric double layer at the particle/aqueous solution interface:
$$ {V}_{\mathrm{T}}={V}_{\mathrm{R}}+{V}_{\mathrm{A}} $$
(1)
$$ {V}_{\mathrm{R}}=\frac{\varepsilon \mathrm{h}\left({\zeta}_1+{\zeta}_2\right)}{4}\kern0.5em \left[\frac{2{\zeta}_1{\zeta}_2}{\zeta_1^2+{\zeta}_2^2}\mathit{\ln}\left(\frac{1+{e}^{-\kappa h}}{1-{e}^{-\kappa h}}\right)+\mathit{\ln}\left(1-{e}^{-2\kappa h}\right)\right] $$
(2)
$$ {V}_{\mathrm{A}}=-\frac{A_{123}\ {a}_1}{6h} $$
(3)
where a1 is the SiO2 radius (0.5 μm) and ζ1, ζ2are the zeta potentials of the SiO2 and magnetite particles, respectively, ε is the medium dielectric constant, κ is the Debye–Huckel reciprocal length parameter, and h is the distance between the surfaces of the interacting particles. A123 is the Hamaker constant, which was calculated as follows:
$$ {A}_{123}=\left(\sqrt{A_{11}}-\sqrt{A_{33}}\right)\left(\sqrt{A_{22}}-\sqrt{A_{33}}\right). $$
(4)

A11 is the Hamaker constant of SiO2, which is equal to 7.9 × 10−20 J, A22 is the Hamaker constant of H2O, which is equal to 3.7 × 10−20 J, and,A33 is the Hamaker constant of Fe3O4, which is equal to 2.1 × 10−19 J [7]. Accordingly, A123 was found to be 2.36 × 10−20 J.

Figure 9 shows VT as a function of the interparticle distance at various pH values. It can be seen that VT was negative at pH 6.5 (process slurry’s pH) indicating that the attractive energy due to van der Waals forces was greater than the repulsive energy due to the negative electric charge at the particle/aqueous solution interface. This explains the heterocoagulation between Fe3O4 and SiO2 during the magnetic concentration process. The mineral particles were stable above pH 9 where VT was higher than 30 VT/kT.
Fig. 9

Total interaction potential energy between Fe3O4–SiO2 particles as a function of the interparticle distance at various pH values

The amount of ultrafine SiO2-bare minerals attached to Fe3O4 was determined by dispersing the particles with Na6(PO3)6, which largely increased the zeta potential of Fe3O4, as shown in Fig. 10. At pH 6.5, the Fe3O4 zeta potential increased negatively from − 12 to − 55 mV with the dispersant. Under these conditions, the repulsive energy was much higher than the attractive energy, so the particles stabilized. Figure 11 shows the particles of Fe3O4 and SiO2 in the dispersed Fe concentrate. As noted, the surfaces of these minerals were completely free of attached ultrafine particles.
Fig. 10

Zeta potentials of SiO2 and Fe3O4 as functions of the pH in the absence and presence of the dispersant Na6(PO3)6 at a concentration of 1 × 10−4 M

Fig. 11

SEM photomicrographs of particles in the Fe concentrate dispersed with Na6(PO3)6

The Fe concentrate was dispersed with Na6(PO3)6 and size fractioned. Then, each size fraction was assayed by SiO2. This SiO2 assay was also performed on the undispersed Fe concentrate. Figure 12 shows the SiO2 distribution as a function of the particle size for the dispersed and undispersed Fe concentrates. As noted, the SiO2 distribution increased by about 6% in the size fraction of − 20 μm for the dispersed Fe concentrate compared to that for the undispersed Fe concentrate. This increase in SiO2 distribution in the finest size fraction was because ultrafine-sized SiO2 detached from the coarse particles of Fe3O4 and passed to the finest size fraction. Then, the amount of SiO2 particles attached to the Fe concentrated by the second drum was determined and was found to represent 5% of the SiO2 in the Fe concentrate.
Fig. 12

SiO2 distribution as a function of size particles in the Fe concentrates dispersed and not dispersed by Na6(PO3)6

3.4 SiO2 Entrainment in Fe3O4 Chains

Magnetite is a ferromagnetic so is magnetized when placed in a magnetic field. The level of Fe3O4 magnetization is established by the magnetic field intensity. This drives the magnetite particles to interact with each other, forming chains where the magnetic interaction force depends on the magnetic field intensity and the size and magnetic susceptibility of the particle [4, 8, 9, 10].

The magnetic field was 0.12 T on the drum surface of the concentration unit used in this work. With this magnetic field on the drum surface, the magnetic field intensity at the bottom of the tank was found to be between 0.005 and 0.006 T [11]. Under these conditions, the magnetite particles entering at the tank bottom formed chains whose lengths and thicknesses increased as the chains traveled to the surface of the magnetic drum, where the magnetic field and magnetic field gradient were very high [4]. During the agglomeration of the magnetite particles, SiO2 particles were trapped between the magnetite chains, as shown in Fig. 13 [9]. The SiO2 entrainment mechanism contributed 75% of the SiO2 in the Fe concentrated by the drum.
Fig. 13

Photomicrographs of Fe3O4 and SiO2 particles at various magnetic fields (a 0.003, b 005, c 0.008, and d 0.011 T) showing the formation of chains of Fe3O4 and the entrainment of SiO2 in Fe3O4 chains in the open circles

4 SiO2 Recovery Mechanisms During Magnetic Drum Concentration

In summary, SiO2 is recovered in the Fe concentrated by the magnetic rotating drum through three mechanisms: (1) mineralogical locking of SiO2 to Fe3O4, (2) heterocoagulation between ultrafine-sized SiO2 particles and Fe3O4 coarse particles due to attractive van der Waals forces and the low negative electric charge at the SiO2 and Fe3O4/aqueous solution interfaces at the pH of the plant slurry, and (3) entrainment of ultrafine-sized SiO2 particles between chains of Fe3O4 agglomerates, which form due to the magnetization of the Fe3O4 particles and the magnetic field in the concentration process. Table 2 presents the contributions of the three mechanism of SiO2 recovery in the Fe concentrated by the magnetic drum, where the SiO2 entrainment mechanism is shown to be the major cause of SiO2 recovery in the magnetic drum concentration process.
Table 2

SiO2 recovery mechanisms and their contributions to the Fe concentration by the magnetic drum

SiO2 recovery mechanism

Contribution, %

SiO2-Fe3O4 heterocoagulation

5

SiO2-Fe3O4 mineral locking

20

SiO2 entrainment in Fe3O4 chains

75

5 Conclusions

Silica (SiO2) is recovered in Fe (Fe3O4) concentrated by a low-intensity-magnetic rotating drum by three mechanisms: (1) mineralogical locking of SiO2 to Fe3O4, (2) heterocoagulation between ultrafine-sized SiO2 particles and Fe3O4 coarse particles and (3) entrainment of ultrafine-sized SiO2 particles between chains of Fe3O4 agglomerates. Heterocoagulation occurs because of the low negative zeta potential values at the Fe3O4 and SiO2/aqueous solution interface of the process slurry pH. Under these conditions, the attractive energy due to van der Waals forces between the minerals is greater than the repulsive energy due to the electric charge at their interfaces. The entrainment mechanism is the major contributor to SiO2 recovery in the drum Fe concentration process.

Notes

Acknowledgements

The authors would like to thank the National Council for Science and Technology, México (CONACyT) for the Ph. D. Fellowship 170585 in Minerals Engineering to Mario Llamas-Bueno and the Ph. D. Fellowships 230133 in Materials Science and Engineering to Mario A. Corona-Arroyo, Autonomous University of San Luis Potosi (UASLP), México.

Compliance with Ethical Standards

Conflict of interest

The authors declare that they have no conflict of interest.

References

  1. 1.
    Wills BA, Napier-Munn TJ (2006) Wills’ mineral processing technology: an introduction to the practical aspects of ore treatment. Butterworth-Heinemann, OxfordGoogle Scholar
  2. 2.
    Yarar B, Dogan ZM (1987) Mineral processing design. Martinus Nijhoff Publishers, BostonCrossRefGoogle Scholar
  3. 3.
    Parker MR (1977) The physics of magnetic separation. Contemp Phys 18(3):279–306CrossRefGoogle Scholar
  4. 4.
    Svoboda J (2004) Magnetic techniques for the treatment of materials. Kluwer Academic Publisher, The NetherlandsGoogle Scholar
  5. 5.
    Sylvester PJ (2012) Use of mineral liberation analyzer (MLA) for mineralogical studies of sediments and sedimentary rocks, Mineralogical Association of Canada Short Course 42Google Scholar
  6. 6.
    Hogg R, Healy TW, Fuerstenau DW (1966) Mutual coagulation of colloidal dispersions. Trans Faraday Soc 62:1638CrossRefGoogle Scholar
  7. 7.
    Israelachvili JN (2011) Intermolecular and surface forces. Academic Press, CambridgeGoogle Scholar
  8. 8.
    García-Martínez HA, Llamas-Bueno M, Song S, López-Valdivieso A (2004) Magnetic flocculation of mineral fines in an external magnetic field. Miner Process Extr Metall Rev 25:67–90CrossRefGoogle Scholar
  9. 9.
    Garcia-Martinez HA, Song S, Lopez-Valdivieso A (2011) In situ observation of quartz particles entrained into magnetite coagulates in a uniform magnetic field. Miner Eng 24:963–966CrossRefGoogle Scholar
  10. 10.
    López-Valdivieso A, Corona-Arroyo MA, Encinas-Oropesa A, García Martínez HA, Aquino- Rosalío CE, Nahmad-Molinari Y (2018) Inhibiting the amine flotation of magnetite through aggregation with uniform low magnetic fields and no chemical depressants. Miner Eng 119:130–136CrossRefGoogle Scholar
  11. 11.
    Llamas-Bueno M (2010) Removal of SiO2 in magnetite concentration by applying external forces in the low-magnetic concentration process (in Spanish), Ph. D. Thesis in Minerals Engineering, Universidad Autonoma de San Luis Potosi, 112 pGoogle Scholar

Copyright information

© The Society for Mining, Metallurgy & Exploration 2018

Authors and Affiliations

  • M. Llamas-Bueno
    • 1
  • A. López-Valdivieso
    • 2
    Email author
  • M. A. Corona-Arroyo
    • 3
  1. 1.Ternium MéxicoSan Nicolás de los GarzaMéxico
  2. 2.Surface Chemistry Lab, Instituto de MetalurgiaUniversidad Autónoma de San Luís PotosíSan Luis PotosíMéxico
  3. 3.División de Ingenierías, Departamento de Minas, Metalurgia y GeologíaUniversidad de GuanajuatoGuanajuatoMéxico

Personalised recommendations