Journal of Scientific & Industrial Research VoL58, June 1999, pp 436-442

Surface Properties of Indian Hematite and Bauxite and Their Coating

Mechanism with Colloidal Magnetite

S Prakash' and B Das Regional Research Laboratory, Bhubaneswar 751 0 1 3

and R Venugopal

Indian School of Mines, Dhanbad 826 004

Received : 05 August 1998; accepted : 1 7 November 1 998

The selective coating of colloidal magnetite on hematite and bauxite for enhancing the magnetic separation techniques which depends on surface and chemical phenomena is studied. The mechanism of colloidal magnetite interaction with hematite and bauxite is investi­gated through electrokinetic and IR studies. The point of zero charge and electrophoretic mobilities of pure hematite, bauxite and colloidal magnetite are determined. The effect of various ions such as PO·\ ' SiO/ and the surfactant such as sodium oleate on surface properties of hematite and bauxite are examined. IR spectroscopy studies indicate that oleate ions are chemically bonded with hematite while it is only physically adsorbed on the surface of bauxite at certain pH. The zeta-potential also correlates the chemical interaction of oleate on hematite. The separation of quartz from hematite and hematite from bauxite is carried out by applying selective colloidal magnetite coating followed by magnetic separation techniques. Hematite is separated in magnetic fraction for both the cases.

Introduction

Hematite is an important raw material widely used for the production of iron and steel. This can be separated from its ore by various methods. In Indian iron ore wash­ing plants, generally washing is being practiced to sepa­rate the associated gangues. However, during the proc­ess of washing around 8 million tonnes of superfines from different washing plants of India with 48-60 per cent Fe are being generated and thrown into the iron ore tailing ponds without any further utility ) . Attempts have been made by different workers to recover the iron values lost by various techniques with l imited success2-4• Similarly, bauxite is also an economic non-metallic mineral used as a raw material for producing alumina and in refrac­tory making. The low grade bauxite contains variety of gangue minerals, viz. Fe-oxides, silicates, and carbon­ates. Amongst them, the separation of iron is problem­atic one, causing difficulty in down stream processes5-7• Roasting fol lowed by leaching is the most general method used for dissolution of bauxite in commercial operations.

Coating and carrier methods for enhancing the mag­netic and flotation separation methods have however re-

* Author for correspondence

ceived little attention8•9. We have been working in our laboratory to enhance the magnetic response of weakly magnetic minerals through selective magnetic coating techniques. In this paper an attempt has been made to understand the surface chemistry and coating mechanism of oleate colloidal magnetite on hematite and bauxite through electrokinetic, IR study and the effect of other ions such as silicate and phosphates for separation in a large scale.

Experimental

The experimental study broadly consist of careful preparation of pure minerals, electrokinetic potential measurements, IR spectrophotometry and wet high in­tensity magnetic separation studies.

Sample

The hematite and bauxite sample, as c lean as possible were used for the experimental studies. The chemical analysis of the samples is given in Table t . The samples used were obtained from Orissa Mining Corporation. Se­lected pieces were crushed and hand ground in a mortar. A portion of the sample was reduced to - 500 11m. After

PRAKASH et al. : SURFACE PROPERTIES OF INDIAN HEMATITE & BAUXITE 437

desliming in laboratory tap water a closed sized fraction of -7S+30�m was separated out by hand sieving. The bauxite sample was washed with O.SN hydrochloric acid to eliminate any trace of iron present in the surface. This was fol lowed by washing and rewashing with distilled water till no trace of chloride ions was detected in rins­ing water. On the other hand the hematite sample was passed in a low intensity magnetic separator to remove any contaminated magnetite in it. Quartz sample was also of mineral origin and supplied by Saroj Mining Corpora­tion, Karnataka. The chemical analysis shows that the sample contains 99.6 per cent Si02• The size and prepa­ration of the material used in the magnetic separation experiment was similar to bauxite and hematite.

Reagents

Reagents used such as perchloric acid, sodium hydrox­ide, sodium silicate, sodium hexa metaphosphate, and potassium nitrate were all AR grade reagents. Fresh so­lutions of sodium oleate (analar grade) having known con­centrations were prepared in situ.

Magnetic separation studies were carried out on the closed sized fractions. A portion of the sample (-7S�m) was further ground and microsieved to obtain -5 �m materia l . Electromagnetic measurement and IR spectroscopy were conducted with the help of -S�m ma­terial .

Methods

Preparation of Oleate Magnetite The colloidal magnetite was prepared by the combi­

nation of I : I ferrous and ferric salts. The salts were dis­solved in distilled water and heated slowly to 70° C. S g of sodium hydroxide dissolved in SO cc of water was added slowly with constant stirring into the warm solu­tion. The precipitated magnetite thus obtained was washed several times, centrifuged to remove salts and sodium hydroxide. The total volume after washing was made up to 1 00 cc. The oleate colloidal magnetite was prepared by the addition of 2 x 10-2M sodium oleate at pH I I and boiling until the magnetite is fully dispersed.

Electrokinetic Measurements

The electrophoretic mobility of the minerals and oleate magnetite were measured by using the zetameter, Rank Brother model, mark-II, UK and the mobility'S were con­verted in to zeta-potentia ls us ing the Helmholtz-

Smoluchowsky equation. Samples for the zetapotential tests were prepared by making a solution of the desired pH by adding the pH regulators, i .e. , HCI04, or NaOH. 0.01 g of the sample was added into the 1 00 ml required solution and it was conditioned for 1 h at the room tem­perature. In all the experiment KN03 (2 x 1 0-5 M) was used as the supporting electrolyte. The mobility of the suspension was determined by adding portion of the so­lution into the given cell connected with two Pt electrodes.

IR Spectroscopy

The IR spectra was recorded by a J ASCO Fourier trans­form IR model S3oo spectrophotometer in 4oo-4000cm­I . The sample used in the spectrophotometer studies was prepared by grinding in an agate motor to a very fine power. One to 2 g of the mineral sample was conditioned at a given pH for I h with the surfactant and the oleate magnetite. After conditioning, the solids were separately dried at SOoC and the FTIR measurements were taken.

Magnetic Separation Studies

The magnetic separation studies were carried out by a wet magnetic separator supplied by Rapid Box Mag sepa­rator, England. The magnetic intensity could be varied with current and grid gaps. The slurry was prepared by adding different proportions of minerals and required quantities of distilled water. The pH of the slurry was adjusted by adding NaOH or HCI04 which was moni­tored throughout the experiment. Oleate magnetite or magnetite fines were added in desired quantity into the suspension. A small quantity of sodium hexa meta phos­phate (0.0 I per cent) was added as dispersant whenever necessary. The mineral mixture was then conditioned for 1 0 min. The slurry was than passed through the magnetic separator slowly. The magnetic fraction was retained on the grid whereas the non-magnetic fraction was collected at the bottom. The magnetic fraction was taken out of the grid by flushing and washing with water in the absence of the magnetic field. The magnetic and non-magnetic fraction thus collected were dried separately and analysed for Fe, Si02, Alp 3' etc. to determine the quality and recovery.

Results and Discussion Electrokinetic Studies

The results of the electrokinetic potential variation of hematite, bauxite, magnetite and oleate magnetite as a function of pH using KN03 as the supporting electrolyte

438 J SCI IND RES VOL.58 JUNE 1 999

> E

�

� c � o Q.

.e .. N

ro�------------�--------------�

• Hemat ite [] Bauxite A Magneiite o Magnetite + Na oleate

11 12

-601L-----�--------p�H------------�

Figure I - Zeta potential of hematite, bauxite, magnetite and magnetite coated with sodium oleate at different pH, (2 x 1 0-3 KNOJ)

60 r---�----------------------.

40

'E 20

c n i l A 1 . 7 x 10-5M • . 3.4 X \05 M o 5.i x lIi5 M

Figure 2 - Zeta potential of bauxite as a function of pH at different concentration of sodium silicate

are given in Figure I . The point of zero charge of hematite is found to be around 7 .2 and that of bauxite is around 8.6. The zero point of charge of hematite reported in the l i terature is at 7.6 to 8.0 (ref. I 0) and that of bauxite is at pH 8.5 to 9.4 (ref. I I ) . For oxide minerals, H+ and OH­ions have since been considered to be potential deter­mining ions. The various concentrations of these ions change the m agn i tude and s ign of electrokinet ic potentials . So these ions are also the potential determin­ing ions for hematite and bauxite, as reported earlier'2 . The pzc of magnetite obtained in the present investiga­tion is around pH 6.8. Above this pH, magnetite has got negative charge. The pzc of synthetic oleate magnetite is about 5 .5 . The synthetic compound resulted a shift in iso-electric point.

Effect of Phosphate and Silicate Ions

Figure 2 and 3 show the electrokinetic potential of bauxite in the presence of sodium hexameta phosphate (SHMP) and si l icate ions. Both these ions play a vital role in the dispersion and thereby the separation of min­erals. It i s seen that the adsorption of sil icate ions or hexameta phosphate ions makes the surface more nega-

oo,---------------·------------� c n i l • 1.7 x 1O:� M A 3.4 X 10_; M o 6.8 x 10 M

12

I Figure 3 - Zeta potential of bauxite as a function of pH at different

concentration of sodium hexametaphosphate

30 ,------------------o---:ni-:-1 ----------, 1 t:. '-' II \O� til

'E 10 ' • 3.4 x 10_5 M o 5.1 X 10 M

L� w " I - 501 L-------

-------------

---------'.

pH

Figure 4 - Zeta potential of hematite as a function of pH and different sodium silicate concentration

> E

�

� c .! 0 0.

" ... .. N

30 20 10 0

-10 -20

-3 -40 - so

4

c n i l • 1 . 7 X 10:;101 o 3.4 . ,0 M -5 Co 5 . 1 X I) M

10 " 12

- 60L-----------_�-----------� pH

Figure 5 - Zeta potential of hematite as a function of pH and sodium hexametaphosphate concent.ration

tively charged bauxite and thereby the particles remain in dispersed phase due to mutual repulsion of the strongly negative charged particles and even at lower pH, the par­ticles remain in suspension. Therefore, either of these ions can be used as a dispersant in the separation of hematite from bauxite by selective magnetic coating.

Similar observations are made with hematite-sodium sil icate (Figure 4) or hematite hexameta phosphate (Fig­ure 5). The electropotential suddenly decreases to nega­tive sides after the addition of the 1 0-5M concentration of sodium silicate. Further addition of the reagent dose

...

PRAKASH e/ al. : SURFACE PROPERTIES OF INDIAN HEMATITE & BAUXITE 439

not change the potential in a significant way. However the mineral possess a constant negati ve potential of 3 1 -35 mv in the pH range 5 to 9 . Only at pH 3.5 the poten­tial drops from + 1 8 mv to around - 1 0 to 1 5 mv. Sodium si licate is therefore one of the most widely used modify­ing reagent in flotation.

The zeta-potential-pH curve for hematite in the pres­ence of d i fferen t concentrat ion of sodium hexa metaphosphate i s shown in Figure 5 . It suggests that the point of zero zeta potential has shifted from pH 6.8 to more on acidic side. From the examination of these curves, it is noted that the addition of hex a metaphosphate to the hematite suspension has produced a change in zeta­potent ial at a l l the pH values s tudied . The hex a metaphosphate has acted as a good dispersing agent of hematite. The interaction between hematite and reagent may also be due to complex formation as phosphates and polyphosphates are good complexing agents for a wide variety of metal cations 1 3.

Effect of Sodium Oleate on the Zeta-potential of Hematite

Figure 6 shows the values of zeta-potential of hematite at different initial concentration of sodium oleate with respect to pH at constant ionic strength. In a selective magnetic coating, surfactants are used along with mag­netite or separately as oleate magnetite in order to con­trol surface charge or better magnetite coatin'g of the sur­face. The adsorption of oleate and magnetite takes place either through the electrostatic interaction or by the chemi­cal interaction between the surfactant and mineral. In the present case, an increase in negative potential starts around 1 O·5M concentration of oleate and continues to increase with further addition of oleate. The increase in negative potential c learly demonstrates the successive oleate adsorption on the surface of the mineral. In the present case the negative oleate ion can only be adsorbed on the mineral surface chemically resulting in the net increase of negative surface charge of the mineral indi­cating specific adsorption .

The zeta potential obtained in the presence of sodium oleate (5x 1 0-5 M) as a function of silicate and meta phos­phate concentration have also been studied. It shows the surface potential is the same -3 1 mv until the 1 0-4 M meta phosphate concentration and further level of these two ions increases in negative zeta potential. As the increase of this potential is marginal, it can be concluded that low effect of these ions might have masked in the presence of oleate similar to apatite and ca1citeI4• It has been stated

that adsorption of sodium oleate on hematite greatly re­duces after treatment with modulus 3.3 sodium silicate. On the other hand, it has been stated that the presence of oleate could possibly cause desorption of sil icates due to high chemical interaction between the cations and the oleate ions.

Results of Infra-red Spectroscopy

The spectra of hematite, hematite conditioned with oleate magnetite at pH 7.2 and the spectra of oleate mag­netite are shown in the Figure 7. Hematite is character­ised by the presence of peaks around 4 1 0-4 1 5 cm- I , 450-470 cm- I and 670-680 cm- I . The absorption bands were detected at 1 559 cm-I , and 1 525 cm-I due to the -C-C and -CH2-CH-C- indicates the formation offerric oleate (Fig­ure 7 III). S imilar peaks detected at 1 590 cml and addi­tional peak at 1 034 cm- I recorded in the case of hematite interacted with oleate magnetite prove the chemical in­teraction of oleate on mineral surface. The peak observed at 1 034 cm- I has been attributed due to -C-O-C cross­linking between the hydrocarbon chains. IR study car­ried out earlier proved that ferrous oleate and aluminum oleate had the characteristic peak at 1 595 cml and 1 6 1 5 cml , respectively l 5. Figure 8 shows the IR spectra of bauxite and bauxite coated with oleate magnetite. Gibbsite the main mineral for bauxite is charatersied by the stretch­ing vibration due to OH group at 36 1 5 cm-I , 3460 cm- I and OH bending vibration at 1 005- 1 035 cm-I and 970-990 cm -I (ref. 1 6) . The IR spectra of bauxite did not alter when bauxite sample was conditioned with oleate mag­netite. There is no chemical interaction between bauxite and oleate species. So the adsorption, if any, may be due to physical adsorption . The mineral conditioned with oleate magnetite was washed several times with distil led water and its IR spectra was recorded again. The IR spec­tra was unchanged and no new spectra appeared.

Result of Magnetic Separation Studies

The separation of hematite from quartz was performed by wet magnetic separation methods by applying the coat­ing techniques. Both oleate magnetite prepared syntheti­cally and finely divided natural magnetite (-5/lm) were used as the coating agents. The best results obtained un­der the optimum conditions are summarized in Table 1 and 2. A synthetic mixture of hematite to quartz ( I : I )

was used. The experiments were conducted at natural pH at which zeta-potential of hematite is also found to be zero and the maximum adsorption of oleate coating takes

440 J SCI IND RES VOL.58 JUNE 1999

Table I - Chemical analysis of hematite and bauxite per cent

Material Fep3 Si02 Alp] CaO MgO LOI

Hematite 97.02 0.92 1 .20 0,06 0.03 0.35 Bauxite 1 . 1 4 1 .85 64.85 0.05 0.03 32.0

Table 2 - Results of magnetic separation of hematite and quartz mixture using colloidal magnetite ( H :Q I : I , pH:7.2, S HMP 0.0 1 per cent)

Details Mag intensity, Weight. Grade, per cent Recovery, per cent

c: :! -o a.

2 _ to N

Magnetic I

Magnetic II

Magnetic JII

Magnetic IV

Non-magnetic

20

k gauss

3.0

5 .0

6.5

7.8

12

-801 �------________________ � pH

per cent

35. 1

47.5

5 1 .6

53.3

46.7

Figure 6 - Zeta potential of hematite as a fuhction of pH and sodium oleate concentration

place at this pH than at the other pH values 17 • The results indicate that a Fe203 grade of 9 1 per cent with more than 97 per cent recovery was obtained by using a magnetic intensity of around 7.8 k gauss. Generally hematite is feebly magnetic in nature and a magnetic intensity of 1 3-1 4 k gauss is required for its effective separation. Since the intensity required is much less than the hematite sepa­ration by a magnetic separation it can be concluded that magnetite response of hematite has been increased by the addition of a small proportion of colloidal magnetite. The addition of sodium oleate as a surfactant with col­loidal magnetite further helped for selective attachment with hematite surface.

Hematite

95.3

94.2

93.5

92.6

1 .5

to u c o � E .. c �

t-

Quartz Hematite

4.7 66.8

6.0 89.3

6.9 96.3

7.7 98.6

98.5 1 .4

4IDl lSOO )000 2500 2000 1500 ' Wave- 'number ( em-I I

Quartz

3.3

5.5

6.9

8.0

92.0

1000 sao

The magnetic separation with natural magnetite also suggested that a Fe203 recovery of 90.3 per cent was

Figure 7 - FTIR spectra of: (I) Hematite, (II) Hematite coated with oleate magnetite, and (JII) Oleate magnetite

�

..

..

,

PRAKASH et al. : SURFACE PROPERTIES OF INDIAN HEMATITE & BAUXITE 441

Table 3 - Results of magnetic separation of hematite-bauxite mixture (H:B 1 : 1 , pH:7.2, SHMP O.OI per cent)

Details Magnetic Intensity, Weight, Product per cent Recovery, per cent k gauss per cent Fep3 Alp] Fep] AIP3

Magnetic I

Magnetic II

Magnetic III

Magnetic IV

Non-magnetic

• " c: " II! E .. c: l! t--J!.

3000

3.0 1 2.2

5.0 44.4

6.5 54. 1

7.8 55.4

44.6

2000 " 1000 Waye number I c m- ' )

Figure 8 - FTIR spectra of: (I) Bauxite and (II) Bauxite coated with oleate magnetite

achieved assaying around 98 per cent Fe203 by using around 6 kg/tonne of magnetite ' 8 . The quantity required is very high compared to the oleate magnetite. However, it has been established that ground magnetite can be used as the coating agents in the system like other minerals8•9•

For the separation of bauxite from hematite which is a very potential impurity in the ore body, a mixture of hematite and bauxite was fed into the magnetic separator using the colloidal oleate magnetite. In this case, sodium hexa metaphosphate and sodium silicate ware used in order to change the surface properties of bauxite. The sign of zeta-potential becomes negative. Therefore the probability of interaction between oleate magnetite with bauxite decreases. As a result, hematite is separated eas­ily from bauxite which has got a chemical interaction with oleate. However, SHMP gave far better results than sodium silicate. The results of the study using 0.0 1 2 g/kg of oleate magnetite and 0.01 per cent SHMP are presented

87.2 1 2.8 2 1 .0 3 .2

92.8 7.2 8 1 .3 6.5

92.8 7.8 " 98.5 8.5

90.7 9.3 99. 1 1 0.5

1 .0 99.0 0.9 89.5

in Table 3. A 99 per cent AIP3 with 90 per cent recovery has bean reported in the non-magnetic fraction. In the absence of SHMP, it has been found that some bauxite was reported in magnetic fraction .

Conclusions

Separation of quartz from hematite or hematite from bauxite can be obtained by using a small fraction of col­loidal magnetite in a low to medium intensity of mag­netic separator. The selective coating and the correspond­ing separation mostly depends on the surface properties of the minerals. Both phosphate and silicate ions play a role in the zetapotential of the minerals. It has also been concluded that colloidal magnetite is chemically bonded with hematite, whereas it is physically adsorbed with the bauxite.

Acknowledgement

The authors are thankful to Prof. H S Ray, Director, Regional Research Laboratory, Bhubaneswar for his keen interest in the study and granting permission to publish this paper.

References

Das B, M Tech Thesis, Beneficiation of Iron Ore Fines of Orissa. ISM, Dhanbad, India. 1 99 1 .

2 Das B , Prakash S, Mohapatra B K, Bhaumik S K & Narasimahan K S, Trans Indian Inst Metals. 44 (5) ( 1 99 1 ) p. 356.

3 Das B , Prakash S, Mohapatra B K, Bhaumik S K & Narasimahan K S, SME. Miner Metall Process, 9 (2) ( 1 992) p. 1 0 1 .

4 Hanumantha Rao K & Narasimahan K S, Int J Miner Proc­ess, 14 ( 1 985) p. 67

5 Das B & Sahoo P K, Ertzmetall, 49 (6) ( 1 996) p. 39 1 .

6 Aglietti E F, Tavani E L, Tedesco & Porto Lopez J M, Hydrometallurgy; 17 ( 1 987) p. 167. "

442 J SCI IND RES VOL.58 JUNE 1 999

7 Indian Bureau of Mines, Report on Beneficiation of Bauxite, Internal Report of IBM, Nagpur, India, 1 990.

8 Parsonage P, Colloid Chern Miner Process, Vol, 12, edited by L S Laskowski and J Raloton (Elsevier, The Netherlands) 1 992 (Development in Mineral Processing).

9 Parsonage P, Trans Inst Miner Metall Sect C, 93 ( 1 984) p. 37.

I O Fuerstenau D W, Pure Appl Chern, 24 ( 1970) p. 1 38.

I I Nonnan L W, SME Mineral Processing Halld Book, Vol.1

(SME, AIMM, New York) 1 985, p. 5-42

1 2 Lai R W M & Fuerstenau D W, Trans AlME, 260 ( 1976) p. 1 04.

1 3 Parsonage P, Int J Miner Process, 20 ( 1987) p. 1 6 1 .

1 4 Hanumantha Rao K H, Britt Maria Antti & Eric Forssberg, Int J Miner Process, 28 ( 1 990) p. 59.

1 5 Leja J , Surface Chemistry of froth Flotation (Plenum Press, New York), 1 982.

1 6 Som S K & Bhattacharya A K , J Ceol Soc India, 45 ( 1 995) p. 427.

1 7 Leslie J, Morgan K P, Ananthapadmanabhan & Somasundaran P, lilt J Miller Process, 18 ( 1 986) p. 1 39.

1 8 Prakash S, Das B, Mohanty J K & Venugopal R, lilt J Miner Process, (accepted).