FLOCCULATION OF IRON ORE FINES USINGMAGNAFLOC·POLYMERS

K.J.Balasubramani*, G.Bhaskar Raju** and P.R.Khangaonkar***

reagent consumption, low settling rate, high turbidityin the supernatant solution and the low shelf life,these substances were mostly replaced by syntheticwater soluble polymers since 1951 [1]. RecentlyMagnafloc and Chemifloc series of flocculants introdu-ced in the market have drawn the attenticm of many~researchers due to their special properties, viz. selec-tivity, low reagent consump.tjon (ppm level), low toxi-city, etc. [2-4]. The present study aims to understand-ing the flocculation behaivour of Kudremukh Iron arefines using Anionic Magnafloc-155 and 1011.

Flocculation of Kudremukh Iron are fines was studi-ed using two commercially available flocculants con-taining anionic polyacrylamide as a basic componentwith different molecular weight. Variables like poly-mer dosage on settling rate and turbidity, pH of thepulp, agitation, pulp density, effect of calcium chlori-

~ de, etc. were systematically examined in and compa-red. Structure of the flocs, IR spectra of iron orefines before and after flocculation were examined inorder to understand the flocculation mechanism. Itwas found that the optimal dosage of flocculantsmechanism. It was found that the optimal dosageof flocculants was around 0.05 kg/t for AnionicMagnafloc-155 and 0.07 kg/t in the case of AnionicMagnafloc-1011. It was also found that the settlingcharacteristics are better with anionic Magnafloc-155compared to ttl~t, of Anionic Magnafloc-1011.

Fines and ultra fines are generated invariably dur-ing mining, milli~g and other metalJurgical opera-tions. Our country is on the threshold of large scalemechanised mining and processing of Iron Ores tomeet both internal demands and export commit-ments. Huge accumulation of fines have led to envi-ronmental problems due to both dust generation and

. water pollution. Since the generation of fines is moreon large scale operations, these fines are to be recove-red and the tailings are to be discharged in an envi-ronmentally acceptable form. Presently these fines.are either ignored or treated empirically by conventio-nal mineral processing techniques.

Flocculation appears to be the most promising tech-nique to solve these problems due to its simplicityon large scale operation and simultaneous water puri-fication for its reuse in the process. In the past natu-ral substances like lime and aluminous earth andnaturai polymers like various starches and Gluewere utilised as sedimentation aids. Due to large

(i) Iron are used in the present investigation wa'sobtained from the mines of Kudremukh, Karna-taka, and the chemical analysis of the sameis shown in Table-1. The lro,n are sample waswet-ground and subjected to seiving to obtain- 400 mesh.

(ii) The details of the f10cculants used are givenin Table-2. The polymer samples were weigh- -ed accurately and added to a known volumeof water and stirred to perpare the solutionsof desired concentrations. Dilute sodium hydro-xide and Hydrochloric acid were used to adjustthe pH of the solution. Analar grade calciumchloride was used wherever necessary.

Total FeSi02

AI203CaOMgOLOI

66.502.751.360.3500404.00

------.:.-------------------:-:-~__=~:__=::::::__:_:==_=_~o=~• Research Fellow, •• Scientist, Natjonal Metallurgical Laboratory Madras Centre, CSIR Madras Complex,Madras-600113 .••• Visiting Professor, School of Mineral Resources Engineering, University Sains Malaysia, 30000~_Ipoh:. Perak,Malaysia.

81. Name of theNo. flocculant

Chemical Nature of Physicalcomposition charge form

Magnafloc-155 Polyacrylamide Anionic Whitebased polymer

Flocculation experiments were done in a graduat-ed cylinder at a slurry consistency of 1% solids.Total volume of the suspension including polymerand other modifiers was made upto the mark of thecylinder by diluting with water. In each experiment,the contents of the cylinder was inverted twentytimes to put the entire system under gentle shakingand was allowed to settle. The settling rate curvewas monitored from the rate of descent of the mud-line (interface between suspension and the supernat-ant liquid) with time. The settling rate values werecalculated from appropriate linear portions of thecurve coinciding with uniform terminal settling veloci-ty of the floes and the same was expressed inem/sec. The flocculation response of the mineral isexpressed as the percentage solids settled in a fixedsettling time. Turbidity measurements of the supernat-ant solution using Nepholo-Turbidity Meter is alsofollowed to measure the degree of flocculation.

Preliminary tests were conducted to see the effectof agitation on mineral flocculation. It was observedthat twenty inversions of the cylinder were enoughto yield optimum flocculation It is also observed thatprolonged and vigorous agitation leads to partial redis-persion and slow rate of settling.

From the Fig.1 it is clear that the steepness ofAnionic Magnafloc-155 is more compared to thnt ofAnionic Magnafloc-1011. It is well known tllnt as tilesteepness of the settling curve increses, settling ratealso increases and greater the steepness, larger isthe relative density and size of the floc (Fig.2). Thesize of the floc at different flocculant dosage was

observed under a microscope. Photographs showthat both the flocculants form bigger floes with increa-sing concentration of the flocculants.lt was also obser-ved that Anionic Magnafloc-155 formed bigg~r flo~s 'than Anionic Magnafloc-1011 (Fig. 3).

The effect of flocculant concentration on settlingand turbidity of the supernatant solution at pH 4.00is shown in Fig.4. The pH of the pulp was so chosenas to correspond to just below the isoelectric pointwhere the Iron are particles are charged positivelY-.Effective flocculation can .be expected by a 1:1.egatlY8:-Iy charged anionic flocculant through charge neuttali-sation mechanism [5]. The, dissolution of the ore wa~ r'found to be negligible at this pH. It is evident -fromFig.4 that the percentage of settling increases withincreasing flocculant concentration. Maximum settl-ing was noticed at 0.05 kg/t for AnionicMagnafloc-155 and 0.07 kg/t in the case of AnionicMagnafloc-1011. Beyond this concentration tharewas no improvement in the settling rate. However,a slight increase in turbidity was observed in supernq-tant solution which may be due to excess flocculantremaining unabsorbed in the solution. Hence ailexact dosage of flocculant is essential to achievethe best settling characteristics. The settling is·- sorapid that within 10 seconds (Anionic Magnafloc-155) and 30 seconds (Anion'ic Magnafloc-1 011) totalsettling of solids. was noticed.

A further set of experiments was conducted tostudy the extent of flocculation as a function of pH(Fig.5). It was observed that the tendency of floccula-tion started decreasing from pH 8.00 onwards. At (pH 10.00 and above, a constant rate of settling wasnoticed. On the other hand, as the pH is made moreacidic, there is only a slight decrease in flocculation.The addition of flocculant improved the floc formationand almost complete settling was noticed within 10seconds for Anionic Magnafloc-155 and 30 secondsin the case of Anionic Magnafloc-1011 in the acidicpH range. Excess dosage of flocculant was foundto have no effect in the basis medium.

Experiments were carried out to obtain flocculationin tile basis medium using these flocculants. Theseexperiments showed that addition of calcium chk>rideplayed a significant role on the settling behaviour(Fig.6). It was found that the minimum amount of

calcium chloride required to achieve flocculation atpH 11.00 was 1.8 kg/t of Iron Ore fines and bestflocculation was noticed only at 5 kg/t. The additionof calcium chloride not only imp~oved the settlingin the alkaline range but also decreased the turbidityof the supernatant solution. Fig. 7 shows the effectof pH on flocculation and calcium chlorideconcentrations.

The sequence of addition of flocculant and calc·ium chloride was found to be an important criterion.From Fig. 8, it is clear that at low concentration off1occulant, better settling and low turbidity could beobtained when calcium chloride was added first.However, when excess amount of flocculant was pre-sent, the flocculation was found to be independentof ttie sequence of addition. But the turbidity wasfound to be more when flocculant was added first.In the alkaline pH, the mineral particles acquire nega-tive charge an.d there exists a strong electricaldouble layer repulsion between the particles andalso b~tween the partical anionic flocculant. The addi-tion of calcium chloride maintains positive charge onthe particle, thus facilitating the absorption of nega-tively charged flocculant. The addition of calcium chlo-ride also reduces the electrical double layer repul-sion between the particles. When an anionic floccul-ant was added to the pulp at this stage, flocculationwas noticed and it could be due to the inter-particlebridging as reported by the earlier workers for anio-nic polymer flocculants [6]. It was also noticed thatin the absence of the flocculant, interparticle bridgingdoes nottai<e place, resulting in poor or nil floccula-tion at this pH.

Wnen calc;:ium 'Chloride was added first, f.ioccula-tion occurred by the above mechanism. But whenthe flocculant is added first, the repulsive forces areincreased since the flocculants also 'possess nega-tive charge. Though the addition of calcium chlorideu~der these conditions decreased the repulsive for-ces, the interparticle bridging was not so significantas noticed in the former case [6].

The flocculated bed depth obtained at different floc-culant concentration was also studied and resultsare given in Fig.9. The bed depth was calculatedafter 30 minutes of the test. From the Figure it isobserved that the bed depth increased initially with

ilocculant concentration attaining a limiting· value andthereafter there was no noticeable difference in beddepth for the two flocculants studied. The amountof flocculant absorbed on the surface of the mineralparticle may be increased upto a certain extent withincreasing amount of flocculant. Due to the hydrationof mineral surface, it swells resulting in increasedfloc size. After certain addition of flocculant concentra-tion, further absorption is not possible and theexcess flocculant remains in the solution and thebed depth formed remains constant. .

I.R.Spectra are recorded using KBr pellet tech-nique to establish the mechanism of polymer absorp-tion on mineral particles. The broad band obtainedfor the flocculated iron ore in the range of 3650-3400CM confirms the hydrogen bond formation. Similarobservation was made by the earlier workers forpolyacrylamide-silica system [7].

The orientation of hydrogen bonding is explainedwith the results obtained form the flocculation respon-se as a funciton of pH. If the hydrogen bonding wasto be established through surface -OH groups, excell-ent flocculation should take place at pH > 6.8 (PZCof iron ore fines), because t1le aissociation of car,boxylate ions will· be favoured with increasing pH.However in the present investigation the flocculationwas found to be very poor in the basis region. Thisobservation indirectly indicates that the hyd~ogen bon-ding is not through surface -OH groups. Since excell-ent flocculation is noticed in acidic pH, it is conclud-ed that hydrogen bonding is most likely through sur-face oxygen species on to which the hydrogen donat-ing groups can be attached. Also, in acid media parti-ally hydrolyzed polycrylamide shall have two typesof hydrogen donating groups :-COOH and-CO-NH2• Here the hydrogen bond via -COOH is stron-ger than via -CO-NH2• due to the greater ele'ctronega-tivity of the oxygen atoms.

The following conclusions were drawn form thethis study :1. Both Anionic Magnafloc-155 and 1011 could be

used as flocculants for Kudremukh Iron Ore finesand that the use of both of them resulted in rapidand complete settling. Amongst the two floccul-

ants studied, Anionic Magnafloc -155 was foundto result in better settling characteristics. The opti-mum dosages were 0.05 kg/t and 0.07 kg/t forAnionic Magnafloc-155 and 1011 respectively.

2. Addition of calcium chloride improved both settlingrate and clarity of the supernatant solution signifi-cantly in alkaline pH range whilst in acidic pHrange, there was no need for addition of calciumchloride to achieve the desired flocculation.

3 The sequence in which calcium chloride and theflocculant added to the pulp was an importantfactor in deciding the efficiency of flocculation.

4. Ttle size and shape of the floc was found to varywittl ttle flocculant concentration and the natureof ttle flocculanL

5. From the experiment results, it was concludedthat effective flocculation was due to chargeneutralisation and inter-particle bridgingmechanisms.

The authors are thankful to Prof. S.Banerjee, Direc-tor, National Metallurgical Laboratory, Jamshedpur,for Ilis valuable suggestion and guidance throughoutthe work.

1. R.f3.Rooth, J.E. Carpenter and HHaertjens, Ind.Min J. (Spl. Issue), 335(1957).

2. NJt!rass, William Alan, S.African _ZA 8605, 609(CI.BOCl):" 25 Mar. 1987, ZA Appl. 85/891, 06Aug, 19U5, 14pp.

3. SA Nicol, Proc. Australas Ins1. Min Metall.No.260, Dec. 1976.

4. Pradip, Prog. Metall. Res FlJndan, Appl. Aspects.,Proc. Int. Cont., 1985 (pub. 1986). 45-50 (Eng.).

5. G.C. Srestry and PSomasundararl, Int .. J. Miner.Process., 6 : 303-320 (1979).

6. K.A. Natarajan and Ramesh Upadhyaya, Tra'ls.Indian Ins1. Metal,s 1986, 39 (6), 627-636.

7. Griot, 0., JA Kitchener, Trans.! Faraday Soc.61, 102li (1965).

K.J.Balasubraman ipost graduated inAnalytical Chemistryfrom the Universityof Madras aQd pre-sently working as aresearch fellow atNML Madras Centrein the field of mineralprocessing especial-Iy on the beneficia-tion of mineral finesby flocculation

Dr. G.Bhaskar Raju is presently working as a Scientistat NML Madras Centre. Post graduated in chemistry fromS.V.University, Tirupathi and obtained doctorate degree l

from the same University on the subject of "Studies onthe flotation of fine particles-electroflotation of chalcopyritefines". He has got 10 yefrs of R&O experience in' thefield of mineral processing and around 12 publications forhis credit.

ProfessorP.8.Khangaonkar-Born in 1935, he obtai-ned his B.E. (Mech.)in 1956, Bf; (Met.) in1957 and ME (Met.)in .1959 from the Col-lege of Engineering,Pune. He obtainedPh.D. (Met.) from Ren-sselaer PolytechnicInst, Troy, USA in (1962. Dr. P.R.K-hangoankar has work-ed in various capaCit-ies in REC, Nagpurand NML Madras Cen-tre and has publish-9d 3 books and morethan 75 papers in lead-ing journals. He hasguided research fel-lows and produced 9Ph.Os so fpr. Present-ly he is working asVisiting Professor.School of MineralResources Enginee-ring, University SainsMalaysia.

CONCENTRATION OF THE

FLOCCULANT kgl t

_ 0·03

30 40 SO 60 70

TIME (sees)

(a)

CONCENTRATION OF THE

FLoCCULANT kg/t

_ 0'02

0-0-0 0·05

••....•••.••• 0'07

D--O-<l 0'10

It-tC_ 0'20

Ev 15•...r.., 12UJr

oo 20 40 60 80 100 120 140

TIME (Sees)

(b)FIG.l TYPICAL SETTLING CURVES OF IRON ORE FINES WITH

(a)A~IONIC MAGNA FLOC-155 (b)ANIONIC MAGNA FLOC-lOll

3'5,3'0

2'5

u

") OIl.~ ........

2·0Eu

UJ•...1'5«

a:C>~ 1·0-'•...•...UJVl

0'5

00

"

--~---------o-o

/

FIG.2 SETTLING RATE OF IRON ORE FINES AS A

FUNCTION OF FLOCCULANT CONCENTRA nON

100

90

80

" 70UJ-'•...

60•...UJVl

UJ 50..,<t•...

40zUJua: 30UJ<l.

20

10

00

\I

\II

~III

I \I I

-rf\ ~I II ,\..~ ..,.~.,

o 70w-':: 60wVl

w 50~<{

~ 40wua:w<l.

SETTLING TIME:

ANIONIC MAGNA FLOC-1SS = 10 sees.ANIONIC MAGNA FLOC-lOll = 3~sees.

>-•...oiiia:::>•... -

a 0o ().(J2 004 006 008 0·10 0·12 0·14 0·16 0·18 {)O20 (}22

AMOUNT OF FLOCCULANT, kg/l

PERCENTAGE SETTLED AND TURBIDITY OF THESUPERNATANT AS A FUNCTION OF FLOCCULANT

CONCENTRATION AT pH 4·0

0-0-0 ANIONIC MAGNA FLOC-155, 005kg/t

~ ANIONIC MAGNA FLOC-lOll, (}Q7 kg/t

SETTLING TIME:

ANIONIC MAGNA FLOC - 155 = lOsees.

ANIONIC MAGNA FLOC- lOll = JOsecs.

D D--4> 1000

, 900,800,,,,700,,

~ 600 :::>•..., , ~/I,' 500 >-•..., , i5, ,400 iii, ,

a:, ,:::>, , •..., ,

300I ,I I, .,

200

"'-s-.~-==i!="V 100

04· 5 6 7 8 9 10 11 12 13

pH OF THE SUSPENSION

FIG.5 PERCENTAGE SETTLED AND TURBIDITY OF THESUPERNATANT AS A FUNCTION OF pH

l00t-;

:~.70

60 t50

>-

500 isiii

400 g;>-

e>-<>-<> ANIONIC MAGNA FLOC -155

0·05 kg/tD-<>-a ANIONIC MAGNA FLOC-lOll

0.07kg/t

pH OF THE PULP = 11·00

SETTLING TIME:

k

~~~"\.II

\\rII'b1

••..•••. -0_..••. ~.::.--o----,o__ ---0-- -0---

ANIONIC MAGNA FLOC-1S5= 10 sees

ANIONIC MAGNA FLOC-lOll = 30secs

a ao 2 4 6 6 10 12 14 16 16 20 22

AMOUNT OF CALCIUM CHLORIDE, kg/t

FIG.6 EFFECT OF CALCIUM CHLORIDE ON FLOCCULATIONOF IRON ORE FINES

100 .~ a B8 B a Il 1000

90 900

80 ~ ANIONIC MAGNA FLOC-155, 0·05 kg/t 800

70 ~ ANIONIC MAGNA FLOC-lOll, 0·07 kg/t 7000

~ AMOUNT OF CALCIUM CHLORIDE ="5 kg/t::J

>- 60 600 ~>-SETTLING TIME:UJ

VI 50>-r\ ANIONIC MAGNA FLOC-155 = 10 Sees. SOOt:ui 0

Cl AN IONIC MAGNA FLOC -1011 = 3D see s iii~ 40 ~\ 400 g;UJu \ \

•...a: J() \ 'l 300UJQ. \ ,

20 \ \, 200\ l..,

10 "---<>-- :"':::'::::8;:.;.-===8=:=8:.-~--o---a 100--0---0-._-0

a 00 4 6 8 9 10 11 12 13

pH' OF THE PULP

FIG.7 EFFECT OF pH ON FLOCCULATION AT FIXEDCALCIUM CHLORIDE AND FLOCCULANT CONCENTRATION

ANIONIC MAGNA FLOC-155,0.05kg!t

ANIONIC MAGNA FLOC -1011,0'07 kg/t

FIRST ADDITION OF FLOCCULANT

SETTLING TIME:

ANIONIC MAGNA FLOC - 155 :: 10 sees·

ANIONIC MAGNA FLOC - lOll = 30secs

pH OF THE PULP = 11·00

100 1000

90 900

60 600

a 70 700UJ ::>....I >-...

60 600 z>-UJ >-<Il >-UJ 50 500 C<.::l iii.<t

400 a:...::>z ...

UJu

30 300a:UJ0-

20 200

10 100

o 0a (}()2 (}()4 0·06 0-06 0-10 0·12 0·14 0;16 0-16 0·20 0-22'

AMOUNT OF THE FLOCCULANT. kg/t

FIG.8 EFFECT OF THE SEQU'ENCE OF ADDLTION OF'FLOCCULANT AND CALCIUM CHLORIDE

o-o-a ANIONIC MAGNA FLOC - '011

pH OF THE PULP = 4'00

PULP DENSITY 10'1,

SETTLING ·TIME = 1&00 ••• c.

FIG.9 VARIATION OF FLOCCULATED BED DEPTH WITH

FLOCCULANT CONCENTRATION

Fig. 3: Effect of flocculant concentration on floc size (X20)at pH 4.00

a) Suspension of iron ore fines before flocculation

(b) Flocculated iron ore using anionic0.025 kg/t (less concentration)

(c) Flocculated iron ore using anionic0.05 kglt (optimum concentration)

(d) Flocculated iron ore using anionic0.1 kg/t (excess concentration)

(e) Flocculated iron ore using anionic0.025 kg/t (less concentration)

(1) Flocculated iron ore using anionic0.07 kglt (optimum concentration)

(g) Flocculated iron ore using anionic0.1 kg/t (excess concentration)

magnafloc-155 of

magnafloc-15S of

magnafloc-1S5 of

magnafloc-1011 of

magnafloc-1011 of

h'fagriafloc-1011 of

13