EXTRACTION OF IRON FROM RED MUD: LOW TEMPERATURE REDUCTION TO MAGNETITE AND MAGNETIC SEPARATION

Sumedh Gostu1*, Brajendra Mishra1, Gerard Martins2

1Department of Material Science and Engineering, Worcester Polytechnic Institute2Department of Metallurgy and Materials Engineering, Colorado School of Mines

Abstract: Red-Mud (Bauxite residue) is a byproduct generated during the Bayers processing of

bauxite. Iron is a major constituent in most of the world’s reserves of bauxite residue (3 billion

tons). The utilization and value addition of iron in red-mud is a critical measure for economically

valorizing the waste and it eases the process of extracting other valuable non-ferrous

components in the waste. A gas based reduction involving a mixture of gases CO (g), CO2(g) and

N2(g) (diluent) is proposed in this paper to convert hematite in red-mud to magnetite. The optimal

conditions for reduction were determined to be temperature of 540 + 10 oC, PCO/PCO2 = 1

(0.070 atm (bar) + 0.001 atm (bar)) for a reaction time of 30 minutes. Magnetic classification of

the reduced magnetite was performed by employing dry (Frantz separator) and wet (Davis tube

separator) means. Approximately 98% of magnetite could be recovered in the magnetic fraction

and the maximum grade of the magnetite achieved was ~60 %. The lower grade of magnetite

was attributed to the presence of nanometer scale entities of agglomerated particulates which

probably comprised of cation lattice-substitutions for Fe+3 by Al+3 and Ti+3 as evidenced in the

STEM images and Mössbauer spectroscopic analyses.

Keywords: Red-mud, Iron extraction, magnetite, Low-temperature reduction, Magnetic-

classification, Nano-size crystallites.

1. Introduction:

The Bayers process for chemical dissolution of bauxite is associated with a selective

precipitation of iron oxide. This has been referred to as red-mud due to the color of its primary

constituent iron (III) oxide (hematite). Iron oxide in conjunction with the Desilication product

(DSP) constitute the red-mud (or Bauxite residue/BR) [1-4]. For every ton of primary aluminum

produced by the Hall Heroult process, 0.8 to 2.5 tons of Bauxite Residue is generated in the

Bayers process depending on the mineralogy of bauxite ore used [5]. The current worldwide

accumulation of BR is estimated at 3 billion tons.

The past half a century saw numerous research focused at utilization of BR. The use of BR as

synthetic soil was one such measure. The high basicity of red-mud would only favor it being a

revegetation measure [6, 7]. The utilization of red-mud in the construction industry has been

researched upon due to the increasing demand of conventional materials for infrastructure.

Bricks, constructional aggregates, roof decking and red-mud based ceramic materials have

been manufactured using red-mud [8, 9]. Cement and geopolymer production from red-mud has

also been intrigued upon but these products cannot compete with the commercially available

conventional products from a quality perspective. Some other applications of red mud such as

waste water treatment for toxic elements [10], cation exchanger [11], filter medium [12],

flocculants [13], a pigment in paints [14] and a filler material for pesticides and insecticides [15]

also prove to be uneconomically, qualitatively unfeasible for commercial utilization with respect

to red-mud’s basicity, complex mineralogy, and presence of fine particulates.

Extracting iron from red-mud could facilitate the utilization of red-mud for some of the above

applications. Metallurgically, the removal of iron could favor the extraction of Aluminum,

Titanium [16] and also concentrate the highly valued rare earth metals to extractable limits [17,

18]. Iron was recovered as a valuable byproduct pig iron by utilizing optimum fluxing and

reductants [19-21]. The slag generated in these processes were utilized to extract Aluminum,

Titanium and other value additions. Enormous research has been done to simultaneously

extract metallic iron, alumina as sodium aluminate [19, 22-26] and aluminum silicates [27] by

employing a soda ash/sodium carbonate and carbon based roasting. Magnetic separation was

performed to separate the metallic iron and sodium aluminate was recovered through water

leaching. Carbothermic reduction of red-mud was performed to produce metallic iron which was

separated by magnetic separation [28-30], yet a clear separation between the ferrous and non-

ferrous components was not achieved in all the reported cases. A similar result was observed

by Regina et al in the separation of paramagnetic Hagg carbide reduced from red-mud. High

Gradient magnetic separation was also tried to separate the iron from red-mud sample, but a lot

of weak magnetic particles were discovered in the magnetic fraction [32]. Hydrothermal leaching

studies to selectively separate iron from the red-mud were also not successful [33-36].

The research reported in this paper proposes a probable techno-economically feasible way to

create value from iron in red-mud. A low temperature reduction process to reduce hematite in

red-mud to produce magnetite and eventually separate it is presented. There has been limited

research on this aspect of iron removal strategy. Magnetite finds its market avidly in computer

drives and loud speakers in view of its ferromagnetic property [37, 38] while the black color of

magnetite makes it a major constituent in paint pigments [39] and the iron content of magnetite

makes it a viable fertilizer for crops [40]. Prior to the reported study, Yanyan Liu et al devised a

method to produce magnetite by anaerobically coroasting pyrite (FeS2) with red-mud. The

elemental sulfur produced helped in this reduction. A wet magnetic separation method was

employed to separate magnetite but the grade of the separated magnetite was not mentioned

by the authors [41]. A similar observation was put forward by the researchers at University of

Missouri Rolla [42].

2. Experimental:

Dewatered red-mud from the Jamaican refinery of Aluminum Corporation of America (ALCOA)

was used for our study. The as received sample contained about 27 % water (L.O.I.). The

sample was dried in a box furnace at 150 oC for 6 hours. The dried material was grounded and

de-agglomerated in a roll crusher with a P80 of 212 µm. A representative sample of red-mud was

generated from the bulk using cone and quarter method. The representative sample was

analyzed for in ICP OES, AAS, XRF, XRD, Mössbauer spectroscopy, TGA, QEMSCAN, SEM

and TEM as a preliminary characterization. The red mud used contained 32 % Iron in the form

of Hematite and goethite, 8 % Aluminum as Aluminum Hydroxides and Alumina, 4 % Titania, 6

% Calcium Carbonate and Desilication product (DSP): Sodium Alumino Silicates.

Red-Mud was reduced in a tube furnace in an atmosphere of CO, CO2 and N2. Stability

diagrams for the reduction system were generated using the thermodynamic data generated

from HSC 5.1. The outlet gases emerging from the tube furnace were passed through an online

CO and CO2 monitor. Mass balance for the process was monitored and % reduction was

measured using Mössbauer spectroscopy.

Magnetite reduced from red-mud using the optimum reduction parameters was subjected to

magnetic separation. Dry and wet magnetic separation routes were tried in Frantz and Davis

tube magnetic separator respectively. Operational parameters related to sample mass flow rate

per magnetic field exposure, magnetic field intensity were varied and optimized. ICP-OES and

Mössbauer spectroscopy were used to analyze the elemental and phase compositions of the

separated fractions.

3. Theoretical Aspects: Phase stability diagram

Thermodynamic analysis of the red-mud Carbon system was constructed utilizing the data from

the HSC chemistry 5.1 software. Gibbs free energy for formation as a function of temperature

was interpreted for different phases plausible during reduction reaction: hematite, magnetite,

wüstite and cementite. The equilibrium thermodynamic diagram is modified to represent relative

partial pressures of Carbon monoxide and Carbon dioxide respectively (reactants for the

reaction system). The reactions among various components in our study are represented in

equations 1-4.

3Fe2O3 (s) + CO(g) = 2Fe3O4(s) + CO2(g) (1)

Fe3O4(s) + CO(g) = 3FeO(s) + CO2(g) (2)

Fe3O4(s) + 6CO(g) = Fe3C(s)+ 5CO2(g) (3)

3FeO(s) + 5CO(g) = Fe3C(s) + 4CO2(g) (4)

The equilibrium representation for the reaction 1 is given by the Van’t Hoffs equation:

ΔGo = -RTlnK (5)

Where R is the universal gas constant and K is the equilibrium constant for the reaction.

For eq 1,

ΔG = -RTln[(aFe3O42

*PCO2)/(aFe2O33*PCO)] (6)

Activities of Fe3O4 and Fe2O3 are assumed to be 1 which leads to equation 7. The values of ΔG

are obtained at various Temperatures by inputting the reaction into the reaction equation tool

box in HSC chemistry 5.1.

ΔG = -RTln(PCO2/PCO) (7)

This exercise is repeated for all the other reaction equilibriums. Then a graph is plotted between

PCO/PCO2 on a decade scale and Log(PCO/PCO2) v/s Temperature and 1/Temperature

respectively in Figure 1. Reduction experiments are conducted in the stability zone of Magnetite

(Fe3O4) varying CO/CO2 ratios and Temperatures.

Figure 1: Stability diagram of Fe-C-O system at total pressure of 0.8 atm.

4. Results and Discussion

Red mud is reduced to magnetite in a carbon based reducing atmosphere. Precursor reduction

using petroleum coke was tried on red mud. Incomplete reduction of hematite in red mud to

magnetite and slower kinetics of reaction were an indication to use of gaseous based reductant

[43]. A mixture of CO, CO2 was used at a constant ratio to fix the reduction potential of the

reaction while N2 was used as a buffer for the experiments. Reductions were performed using

optimized parameters and the reduced magnetite so produced is subjected to magnetic

separation in a dry and wet magnetic separator. The results of some of the experiments are

presented in the following sections:

4.1. Reduction Experiments

Reduction of Jamaican red mud was carried out in the presence of CO(g), CO2(g) and N2(g).The

atmosphere was tailored to lie in the stability zone of magnetite (Figure 1). Experiments were

carried out in an alumina sample boat (50mm (L) * 15mm (W) * 10mm (depth)). After the

reduction, sample is cooled in the natural atmosphere and latter weighed. The reductions were

studied varying CO:CO2 ratio (1:1.5, 1:1, 1:1), temperature (475 oC, 500 oC, 550 oC and 600 oC),

time (10 min, 20 min and 30 min). Elemental, phase analysis and quantification were conducted

0 200 400 600 800 1000 1200 1400 16001

10

100

1000

10000

100000

stability diagram 0.8 atm

Fe2O3 to Fe3O4Fe3O4 to FeOFe3O4 to Fe3CFeO to Fe3CCarbon solutionizing

T (oC)

PCO

/PCO

2

Fe2O3

in ICP, XRD and Mössbauer spectroscopy respectively. 3D contour plots are generated for

three CO:CO2 ratios varying % Magnetite (calculated using Mössbauer spectroscopy) converted

from hematite v/s time and temperature (Figures 2-4).

Figure 2: 3D contour plot (% Magnetite converted v/s time and temperature) at CO:CO2 = 1:1

Figure 3: 3D contour plot (% Magnetite converted v/s time and temperature) at CO:CO2 = 1:1.5

Figure 4: 3D contour plot (% Magnetite converted v/s time and temperature) at CO:CO2 = 1.5:1

For all the CO/CO2 ratios, as temperature increase the % conversion of hematite in red mud to

magnetite increases. A similar trend is observed on increasing the CO/CO2 ratios. Reduction at

550 oC, CO: CO2 = 1:1 and a reduction time of 20 min were chosen as optimum. In addition the

presence of some cementite peaks were seen (3-5 wt %). Reductions were performed using

the optimized parameters obtained using 7 g of sample to generated precursors for next stage

magnetic separation.

4.2. Magnetic Separation

The samples generated under the optimized conditions were subjected to magnetic separation

in a dry and wet magnetic separator. The objective of the experiments was to separate the

magnetite from the non-magnetic fractions of the reduced sample. Dry and wet magnetic

separation were carried out in a Frantz and Davis tube separator respectively. Flow rate and

magnetic field intensity were varied in both the magnetic separation experiments. Elemental

mapping of the magnetic and nonmagnetic portions emanating from the separator. Recovery of

magnetite separated the magnetic fraction v/s grade (purity) of the magnetically separated

magnetite plots were generated for the separation experiments. Locii of all the experiments

conducted in both the magnetic separators are presented in Figure 5 and 6.

40 50 60 70 80 90 1000

20

40

60

80

100

Recovery % (Magnetite)

Grad

e %

(Mag

netit

e)

Figure 5: Grade v/s recovery plot of magnetite recovered in the magnetic fraction in a Frantz

dry magnetic separator.

40 50 60 70 80 90 1000

20

40

60

80

100

RECOVERY % (Magnetite)

GRAD

E%

(Mag

netit

e)

Figure 6: Grade v/s recovery plot of magnetite recovered in the magnetic fraction in a Davis

tube wet magnetic separator.

Figure 5 and Figure 6 present an interesting inference about the magnetic classification. In

some of the experimental parameters for magnetic separation (wet and dry), ~ 90 % recovery of

magnetite in the magnetic fraction is obtained but the grade (purity) of the magnetite classified

remained at 55- 60 %. The presence of very fine particulate agglomerates and/or lack of phase

liberation between various phase constituents in the reduced magnetite could be attributed to

the inefficiency in magnetic classification.

4.3. Phase Liberation

The purpose of this section is to address the phase liberation if present in red mud precursor or

reduced magnetite which was cited as a plausible reason for a decrease in magnetic

classification efficiency in the previous section. Red mud precursor was sonicated in distilled

water and subjected to study under the Transmission electron microscope. This study was

conducted to check if particulate agglomeration was an issue. Figure 7 (a-d) shows the TEM

image of the red mud precursor.

a) b)

c) d)

Figure 7: TEM images of sonicated Jamaican red mud sample

It is observed that the sample consists of numerous nanocrystals of sizes ranging from 40-120

nm. It can be concluded from this study that red mud essentially is composed of small

nanocrystals and the larger particles are essentially composed of small nanoparticles. The

crystal habitat was determined to be orthorhombic, these particles might be inferred as being

lepidocrocite or goethite. It is unknown if the nano particulates are liberated mineralogically.

Mössbauer patterns of red mud, reduced magnetite produced under conditions producing least

and maximum conversion efficiency, pure hematite and pure magnetite are presented in Figure

8.

Figure 8: Mössbauer patterns of select samples

Mag 39: Sample with highest % of residual hematiteMag 36: Sample with highest % of magnetite

Super paramagnetic hematite and Aluminogoethite

3 octahedral, 2 tetrahedral sextets

Cementite

Fe (Fe+3, Fe+2) Fine particles

The presence of additional peaks in the red mud head sample when compared to the pure

hematite sample suggests that the presence of very fine hematite particulates which show the

presence of super paramagnetism and/ or the presence of aluminogoethite. Similarly in the

reduced magnetite sample, the additional sextets present are attributed to the difference in the

neighborhood of the Mössbauer atom Fe57. This might be due to the occlusion of transition

metal atoms (Al and Ti) in the hematite lattice.

There are clear indications of decrease in magnetic classification efficiency owed due to lack of

phase liberation as deciphered through the Mössbauer spectrograms and (or) nano particulate

agglomerations of various phases. A thorough phase classification needs to be conducted for

the nano particulates to ascertain the mineralogical composition.

5. Conclusion

1. The STEM photomicrographs of the sieved red-mud sample indicated the presence of

“particulate” entities. Thus, assigning of a particle size to these entities, which are

clusters of smaller- (nano)size crystallites, may well be an ambiguous (size)

characteristic of this material. Reflecting on how red-mud is produced in the Bayer

Process, this is not entirely surprising. In order to pursue this hypothesis, a sample of

red-mud was sonicated in distilled water to “de-agglomerate” these clusters. Indeed,

STEM photomicrographs reveal the nascent crystallites, which have sizes in the 40 120

nm range. Furthermore, the crystal-habit in some cases may well be associated with

lepidocrocite or goethite, each of which has an orthorhombic crystal-structure.

2. It has been demonstrated that “low” temperature (475oC to 600oC) gas-phase reduction

of hematite in red-mud to magnetite is viable conversion-process that can be achieved

with low partial-pressures of CO(g), and concomitant low partial-pressures of CO2(g). The

low partial-pressures, which are required in order to avoid sooting (CO(g)

disproportionation to C(s) and CO2(g)), require that N2(g) be employed to serve as diluent.

The mass-loss associated with the reduced product (~1012 %) included that associated

with the conversion of hematite to magnetite as well as decomposition of the aluminum-

hydroxide phases also present in the red-mud.

3. Solid-phase reduction-products obtained from the gas-phase reduction of red mud

contained Fe3O4 (56.4 – 80.5 m%), Fe2O3 (0 20 m%), Fe3C (4.8 6.8 m %) and

paramagnetic 2+ and 3+ phases (14 - 22 m%). A paramagnetic (Mössbauer) resonance

is most likely attributable to nano-size iron-oxide phases.

4. The (preliminary) optimal-conditions for gas-phase reduction of the (Jamaican) red-mud

investigated in the research, and reported in this thesis, are: processing temperature of

540oC ± 10C , partial pressures CO(g)and CO2(g) each of 0.070atm (bar) ± 0.001atm.(bar)/

inert diluent-gas: N2(g), for a conversion-time of 30min.

5. Dry and wet magnetic-separation performed on the reduced samples did not achieve a

magnetic (high-iron) fraction and a (low-iron) non-magnetic fraction as the desirable

conversion-product property being sought. This result is most likely attributable to: either,

1) the cation substitution of, primarily, Al3+ and Ti4+/Ti3+ cations in the hydrated-oxide

nanoparticles being converted to magnetite or, 2) nano-size particles of aluminum and

titanium “oxides” occluded within the predominantly “large-particles/clusters” comprising

the precursor red-mud and the subsequent magnetite product.

AcknowledgementsThe authors thank CR3 (Center for Resource Recovery and Recycling) an NSF- IUCRC for

funding this project. The authors also would like to give special commendation to Dr. Don

Williamson, Professor Colorado School of Mines for helping with the Mössbauer

spectroscopy analysis which proved to be detrimental in the quantitative analysis of the

phase constituents.

References

1. Fathi Habbashi, “Textbook of Hydrometallurgy”, 1999; Métallurgie Extractive Québec.

2. Edward, J., Frary, F., Jefferies, Z., “Aluminum and its Production”, 1930; McGraw-Hill

Book Company, Inc.

3. Robert. J. Anderson, “The Metallurgy of Aluminum and Aluminum Alloys”, 1925; Henry

Carey Baird & CO., Inc.

4. Burkin. A.R.,”Production of Aluminum and Alumina”, 1987, Society of Chemical Industry,

John Wiley and Sons.

5. Bauxite Residue Management: Best Practice, World Aluminum, European aluminum

association, April 2013.

6. Liu, Q., Xin, R., Li, C., Xu, C., Yang, J., “Application of red mud as a basic catalyst for

biodiesel production” Journal of Environmental Sciences. 25 (4) (2013), 823-829.

7. Liang, W., Couperthwaite, S, J., Kaur, G., Yan, C., Johnstone, D, W., Millar, G, J., “Effect

of strong acids on red mud structural and fluoride adsorption properties” Journal of

Colloid and Interface Science. 423 (2014), 158-165.

8. Pontikes, Y., Angelpoulos, G., “Bauxite residue in cement and cementitious applications

Current status and a possible way forward” Resources, Conservation and Recycling. 73

(2013), 53-63.

9. Singh, M., Upadhyay. S, N., Prasad, P, M., “Preparation of Iron Rich Cements Using

Red Mud” Cement and Concrete Research. 27 (7), 1037-1046.

10. López, E., Soto, B., Ariaz, M. Nunez, A., Rubinos, D., Barral, M. T. “Adsorbent

Properties of Red Mud and Its Use for Wastewater Treatment”, Wat. Res. 32 (4) (1998)

1314-1322.

11. Liu, Y., Lin, C., Wu, Y. “Characterization of red mud derived from a combined Bayer

Process and bauxite calcination method”, Journal of Hazardous Materials. 146 (2007),

255-261.

12. Ayoub, G., Koopman, B., Pandya, N., “Iron and Aluminum Hydroxy (Oxide) Coated Filter

Media for Low-Concentration Phosphorous Removal”, Water Environment Research. 73

(4) (2001), 478-485.

13. Zouboulis, A., Kydros, K., “Use of Red Mud for Toxic Metals Removal: The Case of

Nickel”, J.Chem. Tech. Biotechnol. 58 (1993), 95-101.

14. Pera, J., Boumaza, R., Ambroise, J., “Development of a pozzalanic pigment from red

mud”, Cement and Concrete Research. 27(10) (1997), 1513-1522.

15. Liu, Y., Zhao, B., Tang, Y., Wan, P., Chen, Y., Lv, Z., “Recycling of iron from red mud by

magnetic separation after co-roasting with pyrite”, Thermochemica Acta. 588 (2014), 11-

15.

16. Logomerac, V, G., “Obtaining Useful Ingredients of Red mud by complex processing”

Rud Metal. Zb. 4 (1976), 313-323.

17. Borra, Chenna Rao, et al. "Leaching of rare earths from bauxite residue (red

mud)." Minerals Engineering 76 (2015): 20-27.

18. Borra, Chenna Rao, et al. "Recovery of rare earths and other valuable metals from

bauxite residue (red mud): a review." Journal of Sustainable Metallurgy 2.4 (2016): 365-

386.

19. Li, X. et al., “Recovery of alumina and ferric oxide from Bayer red mud rich in iron by

reduction sintering” T. Nonferr. Metal. Soc., 19 (5) (2009), 1342-1347

20. Qu, Y. and Lian, B., “Bioleaching of rare earth and radioactive elements from red mud

using Penicillium tricolor RM-10” Bioresour Technol. 136 (2013), 16-23.

21. Jayasankar, K., Ray, P, K., Chaubey, A, K., Padhi, A., Satapathy, B, K., Mukherjee, P,

S., “Production of pig iron from red mud waste fines using thermal plasma technology”

International Journal of Minerals, Metallurgy, and Materials. 19 (8) (2012), 679-684.

22. Qiusheng, Z., Kuangsheng, F., XiaoBin, L., Zhihong, P. and Guihua, L., “Alumina

recovery from red mud with high iron by sintering process” J. Cent. South Univ. Sci.

Technol., 39 (1) (2008), 92-97.

23. Liu, W., Sun, S., Zhang, L., Jahanshahi, S. and Yang, J., “Experimental and simulative

study on phase transformation in Bayer red mud soda-lime roasting system and

recovery of Al, Na and Fe” Miner. Eng., 39 (2012): 213-218.

24. Logomerac, V, G., “Obtaining Useful Ingredients of Red mud by complex processing”

Rud Metal. Zb. 4 (1976), 313-323.

25. Stickney, W.A., Butler, M.O., Mauser, J.E. and Fursman, O.C., “Utilization of red mud

residues from alumina production” Bureau of Mines. U.S. Dept. of Interior, Bureau of

Mines, Washington, D.C., 1970, 32 pp.

26. Zhong, L., Zhnag, Y., Zhang, Yi., “Extraction of alumina and sodium oxide from red mud

by a mild hydro-chemical process” Journal of Hazardous Materials. 172 (2009), 1629-

1634.

27. Bruckard, W.J “Smelting of bauxite residue to form a soluble sodium aluminium silicate

phase to recover alumina and soda “ Min. Proc. Ext. Met. Rev., 119 (1) (2010), 18-26.

28. Bruckard, W.J “Smelting of bauxite residue to form a soluble sodium aluminium silicate

phase to recover alumina and soda “ Min. Proc. Ext. Met. Rev., 119 (1) (2010), 18-26.

29. Yu-hua, G., Jian-jun, G., Hong-jun, X., Kai, Z., Xue-feng, S., “Nuggets Production by

Direct Reduction of High Iron Red Mud” Journal of Steel Research, International. 20 (5)

(2013), 24-27.

30. Jayasankar, K., Ray, P, K., Chaubey, A, K., Padhi, A., Satapathy, B, K., Mukherjee, P,

S., “Production of pig iron from red mud waste fines using thermal plasma technology”

International Journal of Minerals, Metallurgy, and Materials. 19 (8) (2012), 679-684.

31. Caupain, R., “Low-Temperature Gas-Phase Carbidization of iron-bearing constituents in

Red Mud” M.S. Thesis, Colorado School of Mines, 2004.

32. Li, Y., Wang, J., Wang, X., Wang, B. and Luan, Z., “Feasibility study of iron mineral

separation from red mud by high gradient superconducting magnetic separation”

Physica C Supercon., 471(3-4): 91-96

33. Liu, W., Sun, S., Zhang, L., Jahanshahi, S. and Yang, J., “Experimental and simulative

study on phase transformation in Bayer red mud soda-lime roasting system and

recovery of Al, Na and Fe” Miner. Eng., 39 (2012): 213-218.

34. Qu, Y. and Lian, B., “Bioleaching of rare earth and radioactive elements from red mud

using Penicillium tricolor RM-10” Bioresour Technol. 136 (2013), 16-23

35. Agatzini-Leonardou, S., Oustadakis, P., Tsakiridis, P.E. and Markopoulos, C., “Titanium

leaching from red mud by diluted sulfuric acid at atmospheric pressure” J. Hazard.

Mater., 157(2-3) (2008), 579- 586.

36. Zhong, L., Zhnag, Y., Zhang, Yi., “Extraction of alumina and sodium oxide from red mud

by a mild hydro-chemical process” Journal of Hazardous Materials. 172 (2009), 1629-

1634.

37. Tang, Jing, et al. "Magnetite Fe3O4 nanocrystals: spectroscopic observation of aqueous

oxidation kinetics." The Journal of Physical Chemistry B 107.30 (2003): 7501-7506.

38. Raj, K., and R. Moskowitz. "Commercial applications of ferrofluids." Journal of

Magnetism and Magnetic Materials85.1-3 (1990): 233-245.

39. Legodi, M. A., and D. De Waal. "The preparation of magnetite, goethite, hematite and

maghemite of pigment quality from mill scale iron waste." Dyes and Pigments 74.1

(2007): 161-168.

40. Elfeky, Souad A., et al. "Effect of magnetite Nano-Fertilizer on Growth and yield of

Ocimum basilicum L." Int. J. Indigenous Med. Plants 46.3 (2013): 1286-1293.

41. Liu, Y., Zhao, B., Tang, Y., Wan, P., Chen, Y., Lv, Z., “Recycling of iron from red mud by

magnetic separation after co-roasting with pyrite”, Thermochemica Acta. 588 (2014), 11-

15.

42. Xiang, Q., Liang, X., Schlesinger, M., Watson, J., “Low-Temperature Reduction of Ferric

Iron in Red Mud”, TMS (The Minerals, Metals and Materials Society), 2001.

43. Gostu, Sumedh, Brajendra Mishra, and Gerard P. Martins. "Low Temperature Reduction

of Hematite in Red-Mud to Magnetite." Light Metals 2017. Springer, Cham, 2017. 67-73.