Flocculation characteristics of kaolin

Flocculation characteristics of kaolin

Applied Clay Science, 6 (1992) 383-393 383 Elsevier Science Publishers B.V., AmsterdAm Flocculation characteristics of kaolin F.F.O. Orumwense, C.A...

539KB Sizes 1 Downloads 63 Views

Applied Clay Science, 6 (1992) 383-393

383

Elsevier Science Publishers B.V., AmsterdAm

Flocculation characteristics of kaolin F.F.O. Orumwense, C.A. Eligwe 1 and J.U. Ejiofor 2 Metallurgical and Materials Engineering Department, Collegeof Engineering and Technology, Anambra State Universityof Technology, Enugu, Nigeria (Received June 11, 1991; accepted after revision November 25, 1991 )

ABSTRACT Orumwense, F.F.O., Eligwe, C.A. and Ejiofor, J.U., 1992. Flocculation characteristics of kaolin. Appl. Clay Sci., 6: 383-393. The effect of metal cations on the flocculation behaviour of kaolin was investigated. The effects of pH, concentration of cations (Mg 2+ and AI3+ ), and the presence of these cations in causticised cassava starch on the settling rate and percent reduction in sediment density were studied. Results show that the presence of high concentrations of divalent (Mg2+ ) and trivalent (AI3+ ) ions, and the presence of these cations in causticised cassava starch significantly contributed to the existence of low settling rates at high supernatant clarity often observed in some clay slurry treatment systems by floeculation and sedimentation method. It was also established that for efficient flocculation, the systems must be within the alkaline pH range (pH> 10). The study further revealed that in some flocculation systems, fast settling corresponds to high sediment volume whereas in others, the opposite applies. It is therefore concluded that this phenomenon should be taken into consideration for proper design and operation of such treatment facilities.

INTRODUCTION

In the production and subsequent processing of kaolin, the state of aggregation of the clay particles is a parameter which needs careful control to achieve maximum efficiency and to obtain optimum results. Thus, the addition of flocculants or deflocculants to the slurry becomes necessary and a small variation in the concentration of these materials can have remarkable effects on the processing. For instance, the use of an excessive amount of deflocculant during production and refining is a financial waste which may be compounded at a later stage by the need to add a corresponding excess of flocculant to optimise solid-liquid separation. Furthermore, the nature of some of Tresent address: Federal University of Technology, Owerri, Imo State, Nigeria. 2present address: Projects Development Institute, PRODA, Enugu, Nigeria. Correspondence to: F.F.O. Orumwense, Anambra State University, Department of Metallurgical and Materials Engineering, P.M.B. 01660, Enugn, Nigeria.

0169-1317/92/$05.00 © 1992 Elsevier Science Publishers B.V. All rights reserved.

384

F F . O ORUMWENSE ET AI

these additives requires their concentration to be kept to a minimum to avoid corrosion of the plant and machinery and to improve product quality. Two groups of flocculants used in most washing plants are the non-polymeric and polymeric flocculants. The former are inorganic electrolytes such as lime, calcium chloride, magnesium chloride, alum and other aluminium salts (e.g. aluminium chloride or nitrate) and ferric sulphate. The later are either synthetic organic polyelectrolytes such as those sold under the trade names superfloc, separan and Nalco or natural polymers such as starches, guargum, tannins and sodium aliginate (Matheson and Mackenzie, 1962; Bratby, 1980 ). These natural polymers have the advantage of being non-toxic, whereas some synthetic polymers may be toxic. Natural polymeric flocculants have found use in the mining industry as filtration aids (Mackenzie, 1980), selective flocculants in ore beneficiation (Read and Hollick, 1976; Shaning and Attia, 1987) and selective depressants for vien minerals in ore flotation (Read and Hollick, 1976 ). These processes, including bulk flocculation, involve the initial adsorption of the polymers at the solid-liquid interface. Recently, modified cassava starch was successfully used (Orumwense et al., 1990) to study the adsorption and flocculation response of coal-washery effluent particles and the flocculation behaviour of a kaolinitic clay suspension. Small amounts of oxidized cassava starch were found to be needed for effective adsorption and subsequent flocculation of coal-washery effluent slurries at all pH's except in the highly alkaline range (pH > 8 ), and adsorption was found to have mainly resulted from electrostatic forces. Charged oxidized cassava starch was found to be a better flocculant than uncharged oxidized cassava starch for the kaolinitic clay suspension, and flocculation process was due to both hydrogen bonding and electrostatic interaction. Current practice indicates that synthetic polymers are used in preference to the inorganic electrolytes because the synthetic polymers are cheaper and small amounts of these synthetic polymers are very efficient. Cationic polymers are effective flocculants for the particles found in coal-washery effluent slurries ("blackwater"-containing suspended solids of fine coal, clay particles, etc. ). These cationic polymers are not very attractive because they produce small floc size and consequently low settling rates lead to low thickener capacity. Anionic polymers alone are not used in the flocculation of coal-washery slurties because at neutral pH, the majority of the particles found in "blackwater" are negatively charged and electrostatic repulsion hinders the adsorption of the polymeric anions. Indeed, it has been reported that anionic polymers are poor flocculants for "pure" clays in pure water (Hogg, 1980). Anionic polymers in the presence of inorganic cations such as Mg 2+, A13* and Fe 3+ are quite effective flocculants. Also, starch-based and non-ionic synthetic polymers perform best in the presence of metal ions. Furthermore, in areas where synthetic polymers are not readily available, metal cations are still being exclusively used as flocculants. In general, metal cations are always present in

FLOCCULATIONCHARACTERISTICSOF KAOLIN

385

clay slurry treatment systems either added deliberately during flocculation or invariably present in the particulates found in the slurries. The work reported here was designed to determine how such factors as pH, cation concentrations, the presence of metal ions in causticised cassava starch affect the floeculation response of clay slurries. EXPERIMENTAL PROCEDURE

Materials The kaolinitic clay used in this study was procured from the clay deposit at Nsu in Imo state of Nigeria in lumps ranging from 5 to 20 ram. The clay lumps were crushed and ground to fine size and lawned through 200 mesh (75 am), British standard sieve. Mineralogical analysis of the clay showed the major constituents to be kaolinite (69%), quartz sand (19%), feldspar (8%) and CaCo3 (1.0%). The flocculants used in the tests are salts of divalent and trivalent metals, and a polymer. The salts are magnesium chloride hexahydrate (MgCI2-6H/O) and aluminium nitrate nanohydrate AI(No3)s-9H20. The polymer is 1% causticised (with NaOH) cassava starch. Deionised water was used in the preparation of the reagent solutions. Dilute solutions of NaOH and HC1 were used for pH control.

Method Four grams of the clay material was dissolved in 200 ml of deionised water to give a slurry containing 2% solids by weight. Each slurry was transferred to a 500 ml beaker and the pH was then adjusted to the desired value ranging from pH 7.5 to 12. The experiment was carried out in three stages: ( 1 ) without the addition of any flocculant so that the effectiveness of the flocculant could be judged; (2) addition of electrolyte coagulants in the concentration range of 10-6 to

10-1 mol/1; (3) introduction of the 1% cansticised cassava starch together with the electrolytes in the concentration range of 10-5 to 10-2 mol/1. Each sample was poured back and forth twelve times between two 500 ml beakers. This mixing method was found to yield reproducible results. The sample was then poured into a 500 cm 3 cylinder to which a millimeter scale had been attached on the side. The settling curve of the floes was then monitored by taking the height of the mud line (interface between the supernatant liquid and the suspension) at five minute intervals until constant readings started to occur. A curve of mud line height versus time was plotted for each test and the settling rate values were calculated from the approximately linear

386

F.F.O. ORUMWENSE

ET AL

portion of the curve coinciding with the uniform terminal velocity of the flocs. After taking the mud line readings, the cylinder was left undisturbed for 24 h after which the sediment volume for the particular test was recorded. Flocculant performance was also monitored by measurement of the sediment density by the method of Michaels (1954) as well as by visual observation of the supernatant. The percent reduction in sediment density with respect to supernatant clarity was then used as a performance index for the flocculants: high percent reduction in sediment density corresponding with low-turbidity supernatant. The electrolyte solutions were "aged" for 48 h before being used in the flocculation tests to standardize the effect of metal cation hydrolysis (Bratby, 1980 ). All tests were conducted at room temperature (25 + 3 °C ). RESULTS AND DISCUSSION

The results based on settling rate and percent reduction in sediment density are presented in Figs. 1-7. Figure 1 shows typical settling curves of the flocculation tests. The floes initially settle at a very fast rate and gradually level up with time. This behaviour is due to the gravitational force acting on individual floes which forces it to settle freely when the floes have large separation distances from each other. As the settling proceeds, the inter-floe distances become reduced (i.e., pulp density increases), and the tendency of the flocs hindering themselves from free settling increases. The results of the effect of pH on flocculation behaviour of the clay slurries [ Mg2. ],mol/I

a--o.s.162 I - - o-3.~62 5

,c 2

0

0

I

I

I

I

I

I

I

5

10

15

20

25

30

35

Time(min)

Fig. 1. Typical settling rate curve with MgCI2- 6H20.

40

PLOCCULATIONCHARACTERISTICSOF KAOLIN

387

are depicted in Figs. 2-4. Preliminary tests showed that the clay particles were not flocculated in the acidic pH region but a good flocculation performance of the flocculants was observed in the alkaline region. The effect of pH on the flocculation performance of a 4 × 10 -2 molar solution of Mg 2+ ions is given in Fig. 2. As can be observed, the percent reduction in sediment density increases non-linearly with pH, reaching a maximum value of about 49% at pH 11.5 before it starts to decrease. This implies that optimum supernatant clarity occurs between pH 11 and 12. It is also noticeable in Fig. 2 that settling rate on the other hand decreases with pH and that maximum percent reduction in sediment density occurs at low settling rates. Figure 3 shows the extent of flocculation with A13 + ions as a function of pH. As can be seen, the presence of 7.5 × 10- 3 molar A13 + ions increases the percent reduction in sediment density linearly with pH. Also, it can be observed that the settling rate decreases with increasing pH and at about pH 11.5, one notices a sharp drop in the settling rate to low values, thus indicating again that high supernatant clarity does not correspond to high settling rate. The flocculation response of starch in the presence of 10-2 molar M g 2+ and Al3+ ions as a function of pH is shown in Fig. 4. As can be seen from the figure, the percent reduction in sediment density increases non-linearly with increasing pH, reaching their maxima (65 and 70%, respectively ) at pH 1 I. 5, after which it starts to decrease. This implies that optimum supernatant clarity occurs around pH 11 to 12. Figure 4 further shows that the lowest settling rates (0.9 and 0.6 c m / m i n ) were recorded at pH I 1.5, and that maximum percent reduction in sediment density occurs at low settling rates. This indi-

Concentration of M2°= 4.102 mol/I

60

5o.~ ~4

~o~

3

~

30._E

2

20._c

E

g

0

8

I

I

I

I

I

9

10

11

12

13

o 14

PH

Fig. 2. Effect of p H on flocculation with MgC12" 6H20.

388

F.FO

Concent~atior, of ~.13":75.10~3 mot.q

./

2.5

E

E'I' AI,

60

2.6

c

ORUMWENSE

50~ >.

2"4

40 ca

3o~

* 2,3 r~

2O c

~,2.2

._c

~n2. I n~

0

2.0 8

10

9

pH

11

12

14

13

Fig. 3. Effect of pH on flocculation with A1 (NO3)3" 9H20. 1.5

95

Conc =40ppm AI3"Conc =102mol/[

Starch

Starch Cone =40ppm Mg2° Cone = 102mot/l

~0

[b]

8o 70~m

1"0 c

50

~o~ 0-5

.= 40

!3o§ I

I

I

B

9

10

pH

I

I

1

11

12

13

7

8

9

pH

10

11

12

13

Fig. 4. Flocculation response of starch in the presence of metal ions as a function of pH. (a) Mg 2+. (b) A13+.

cates that very high supernatant clarity can only be obtained at the expense of some reduction in settling rate. With reference to Figs. 2-4, the increase in percent reduction in sediment density observed as pH increases (below the isodectric points for Mg (OH)2 and AI (OH)3, which occurred at about pH 11.5 and 9.3, respectively could be attributed to the increase in the amount of adsorbable positive polynuclear species, and hydrogen b r i ~ n g has been identified to be responsible for the

FLOCCULATION CHARACTERISTICS OF KAOLIN

389

adsorption of these metal ion complexes onto the mineral particles. The adsorption of the oppositely charged complexes on the mineral surface results in the reduction of the electrokineticpotential of the mineral and the particles have a tendency to coagulate and flocculate spontaneously as they approach their isoelectric points (IEP). The decrease observed in percent reduction in sediment density for Mg2+ ions above the isoelectric point is due to the formation of anionic polynuclear complexes rather than the cationic ones leading to electrostatic repulsion and thus restabilising the system. The continued increase in percent reduction in sediment density observed well above AI(OH)3 isoelectric points (Figs. 3 and 4b) suggests that in the highly alkaline region, the electrostatic forces are not sufficient to overcome the destabilising factors existing in this region. This is an indication that the A13+ ion is a better flocculant than the Mg2+ ion. Figures 5 and 6 show the effect of Mg2+ and AI3+ ion concentrations on flocculation at pH 11.5. It is to be noted, however, that the flocculation tests performed with cation concentrations below those reported (Figs. 5-7), in the presence or absence of starch did not lead to flocculation or when flocculation occurred, no clear mud line was observed and so the results are not reported. Figure 5 shows that both settling rate and percent reduction in sediment density increase with increasing Mg2+ ion concentration. This indicates that at this pH value, increase in electrolyte concentration results in a corresponding increase in supcrnatant clarity as well as generation of large fast settling floes. However, at about 5 X 10-2 molar Mg2+ ion concentration, 2.7

1

70

pH =11'5

60

~" 2"6f 50 "~

E2.4

4o

E 3O N

~

2.2

2.0

30 40 50 60 70 80 Concentration of Mg2. ion (rnol/I ~I(]3 ) Fig. 5. EffeCt of Mg 2+ ion concentration on flocculatJon. 0

10

20

90

0

100 105

390

i,~,F tL O R U M W E N S E

E T A[

the settling rate reaches a m a x i m u m value of about 2.6 c m / m i n , after which a gradual decrease sets in. The percent reduction in sediment density on the other hand increases up to about 38% at about 5 × 10- 3 molar Mg= + ions and then starts to decrease. This again indicates that there is better flocculation at low settling rates. In the tests involving A13+ ions (Fig. 6), the percent reduction in sediment density increases with increasing A13+ ion concentration up to about 7.5 X 10 -3 molar. Above this value, a gradual decrease sets in. Also from Fig. 6, it can be seen that the settling rate curve is S-shaped. This shows that settling rate increases with increasing A13+ ion concentration until it reaches a m a x i m u m value of about 2.45 c m / m i n at 7.5 × 10 -3 molar A13+ ions. Beyond this value, it starts to drop to a m i n i m u m value of about 2.35 c m / m i n at an AI 3+ ion concentration of 8.8 × 10-3 molar. Above this value, the settling rate starts to decrease, thus indicating the formation of large fast settling flocs once more. The increase in percent reduction in sediment density observed with increase in ion concentration (Figs. 5 and 6 ) could be as a result of the presence of high concentration, and the adsorption of the metal ion complexes at the mineral surface. The decrease observed may be due to charge reversal on the colloidal surface (i.e., a reversal of the sign of the zeta potential in the double layer) leading to repulsion and restabilisation of the system. The results of the tests with starch in the presence of 10 -2 molar Mg 2+ and AI 2+ ions are shown in Fig. 7. From the figure it can be observed that at low polymer dosage, the settling rate decreases and shoots up again at higher concentration values. However, greater reduction in settling rate was recorded with the presence of A13+ ions in starch than with Mg 2+ ions. The percent 50

2,6

pH 11-5

so .~ 2,4

D

E

30 ,~ 20 ~

o~ 2,2

lo }-g rY

2.0

I

5

I

i

6

7

Concenlrotion

of

i

J

8

9

AI 3. ion(mol/1.10

I

-2 10

0

11

.)

Fig. 6. Effect o f A P ÷ ion concentration on flocculation.

391

FLOCCULATION CHARACTERISTICS OF KAOLIN

1,6 1.5

120 Idg ° ion

AI 3" ion Conc=l(~ 2 rnol/I

Cone: ~ 2 r n o l / I

pH:11.5

pH=11.5 [

110

bJ 100

1.0

90

c E

80

w(u 0"5 60 g

5O ~

30

I

I

I

50

70

90

Starch

I

116

Concentration, p pm.

130

i

I

I

150160 30

50

70

Starch

I

90

110

i

130

,

~0

150160

Concentration, ppm.

Fig. 7. Effect of starch concentration on flocculation in the presence of metal ions. (a) Mg 2+.

(b) AI3+.

reduction in sediment density curves are very similar. It increases with increase in starch concentration. This is probably due to the bridging action of the polymer chains on the polynuclear species and on the mineral particles leading to an increase in sediment volume. The settling rate and reduction in sediment density curves (Figs. 2 - 7 ) show that in some cases fast settling correlates with high percent reduction in sediment density (e.g. Figs. 6 and 7), whereas in others (e.g. Figs. 2-5 and 7), high supernatant clarity can only be achieved at low settling rates. This could, however, be due to differences between "free-settling" and "hindered-settling" regimes (Payatakes, 1975 ) as well as to the packing nature of the flocs. Large flocs cannot pack as closely as small ones so they produce more porous sediments and larger volumes. CONCLUSIONS

The present study has lead to a number of findings: ( 1 ) Flocculation of the kaolin slurries from Nsu in the absence of di- and trivalent cations is not possible at all pH values. For flocculation to occur, high concentrations of di- and trivalent cations (Mg 2+ and A13+ ) must be present. For efficient flocculation, the system must be within the alkaline pH region ( p H > 10). (2) Flocculation is markedly improved using A13+ or Mg 2+ ions with caus-

392

F.I,' O. O R U M W E N S E ET AI,

ticised cassava starch. The flocculability is increased at a very slow rate as starch concentration increases. (3) In some flocculation systems, fast settling corresponds to high sediment volume whereas in others, the opposite applies. In conclusion therefore, it is important to take this phenomenon into consideration during the design and subsequent operation of clay slurry treatment facilities. ACKNOWLEDGEMENTS

The authors wish to thank Dr. B.C. Egboka and Dr. Orajiaka both of Anambra State University of Technology, Enugu, for their interest and assistance throughout the duration of this work. APPENDIX A - - CALCULATION OF PERCENT REDUCTION IN SEDIMENT DENSITY

Percent reduction in sediment density original sediment density- test sample sediment density (without reagent) (with reagent ) original sediment density (without reagent)

where the sediment density (determined by leaving the sample in the cylinder undisturbed for 24 h to obtain the sediment volume) is calculated as: amount of solid in sample Sediment density = volume of sediment after 24 h

REFERENCES Bratby, J., 1980. Coagulation and Flocculation. Uplands Press, Croydon, pp. 136-171. Hogg, R., 1980. In: P. Somasundaran (Editor), Fine Particles Processing. Vol. 2. Am. Inst. Min. Metall. Pet. Eng., Inc., New York, N.Y., pp. 990-999. Mackenzie, J.W.M., 1980. Guar-based reagents. Eng. Min. J., 181 ( 10): 80-87. Matheson, G.H. and Mackenzie, J.W.M., 1962. Flocculation and thickening Coal-washery refuse pulps. Coal Age, 67( 12): 94-100. Michaels, A.S., 1954. Aggregation of suspensions by polyelectrolytes. Ind. Eng. Chem., 46 (7): 1485-1490. Orumwense, F.F.O., Okorrie, B.A. and Eligwe, C.A., 1990. Adsorption studies of modified cassava starch on coal-washery effluent particles in relation to flocculation. Scand. J. Metall., 19: 153-158. Orumwense, F.F.O., Eligwe, C.A. and Mbah, C.N., 1990. Flocculation behaviour of kaolinitic clay suspension using oxidized cassava starch. Appl. Clay Sci., 5: 295-306. Payatakes, A.C., 1975. In: F.M. Tiller (Editor), Theory and Practice of Solid-Liquid Separation. Univ. Houston Press, Houston, pp. 135-213.

FLOCCULATIONCHARACTERISTICSOFKAOLIN

393

Read, A.D. and Hollick, C.T., 1976. Selective flocculation techniques for recovery of fine particles. Min. Sci. Eng., 8:202-213. Shaning, Y. and Atria, Y.A., 1987. In: Y.A. Attia (Editor), Flocculation in Biotechnology and Separation Systems. Elsevier, Amsterdam, pp. 601-637.