Powder Technology, 38 (1984) 267 - 273 267

Mechanisms of Grinding Modification by Chemical Additives: OrganicReagents

.H. EL-SHALL

Mineral Proceuing Engineering, Montana College of Mineral Science and Technology, Butte, MT 5970t (U.s.A

and P. SOMASUNDARAN

Henry Krumb School of Mine" Columbia University, New York, NY 10027 (U.S.A.)

(Received August 20, 1982; in revised form August I, 1983)

SUMMARY

The effect of dodecylammonium chlorideon properties of quartz slurries that arerelevant to grinding, such as pulp fluidity,flocculation and dispersion, and primarybreakage, have been investigated in this study.These tests were conducted under chemicalconditions (pH, ionic strength, additiveconcentration, etc.) similar to those used inwet ball milling tests. The data obtained werecorrelated with the grinding results. Aminewas found to improve grinding, pulp fluidity,and primary breakage of quartz suspensions,especially in the alkaline pH range. Thesebeneficial effects are attributed to the forma-tion of highly surface-active amineaminiumcomplexes in this pH range. In contrast to theabove, amine decreased the grinding rate inthe acidic pH range owing to increasedflocculation of quartz fines in the presence ofamine in this pH range. These results showthat grinding of quartz can be improved byamine if other conditions such as pH arecontrolled. Also, the results indicate that theeffect of chemical additives on grindingcannot be attributed to a single mechanism,but that the cumulative effect on differentpulp properties has to be taken into con-sideration. Examination of the effect of allproperties has led to a better understandingof the mechanisms responsible for the ob-served grinding aid effects in the presentstudy.

for the observed effects of chemical additiveson grinding: the first based on the reductionof surface-free energy of solids induced by theadsorption of surface active agents [1], thesecond related to changes in surface hardnessdue to the influence of the adsorbing specieson the mobility of near-surface dislocations[2]. Based on the results of a research pro-gram studying the effect of fluidity modifierson the performance of batch and industrialgrinding mills, other workers [3 - 13] haveadopted a third mechanism. The proposedmechanism has emerged from correlation ofthe grinding results with the rheologicalbehavior of slurries of similar solid content.Such correlations indicate that certain water-soluble chemicals can improve the fluidity ofhighly concentrated solid suspensions, leadingto higher throughputs or finer products atconstant feed rates.

The above discussion suggests that theeffect of chemical additives on the grindingprocess has been attributed to individualmechanisms. However, grinding is an integralprocess involving several simultaneous sub-processes such as transport of particles to andfrom the grinding zone, where they aresubjected to different kinds of stresses.Evidently, such subprocesses depend on pulpproperties such as fluidity, state of floccula-tion or dispersion, and mechanical propertiesof particles. Therefore, the individual as wellas the cumulative environmental effects onthe pulp properties should be consideredwhen formulating mechanisms controlling theinfluence of chemical additives on grinding.

In this study, the results of adding dodecyl-ammonium chloride during quartz wet ballmilling are correlated with its effects on basicproperties of the system such as pulp fluidity,

INTRODUCTION

Past investigators have considered twomajor mechanisms to be mainly responsible

0032-5910/84/$3.00 @ Elsevier Sequoia/Printed in The Netherlands

268 "account for the iron that is released from themill during grinding. After tumbling thesuspension for 10 min, a 6 ml sample waspoured into the instrument holding tube tomeasure the percentage light transmitted. Allthe measurements were made at a wavelengthof 620 nm. In each case, the instrument wascalibrated with water for a reading of 100%transmission.

flocculation and dispersion, and crack initia-tion and propagation (primary breakage).

EXPERIMENTAL

Material and chemicalsA highly crystalline quartz sample of 99.9%

purity (Hot Springs, Arkansas variety) pur-chased from Ward's Natural Science Establish-ment was used in this study.

The amine was dodecylammonium chloride(DDACI) of practical grade purchased fromEastman Kodak. pH adjustments were madeusing reagent grade NaOH or HC!. Reagentgrade NaCI was used to adjust the ionicstrength. Fe(NO])] or FeCI] crystals pur-chased from Fisher Scientific Company wereused in the basic studies to simulate thedissolved iron introduced into the mill duringgrinding. Triple-distilled water was used forthe preparation of all solutions.

Equipment and methodsIn this study, the following experimen~

were conducted:

3. Rheological studies (torque measure-ments)These experimen~ were performed under

selected conditions to identify the effect ofthe above additives on the flow behavior ofpulps at high solid concentrations. Theequipment and experimental conditions forthe measuremen~ are summarized below.Driver and control: Master servodyne (Cole-Palmer, Chicago); constant speed and torquecontrol unit model 4424; digital millivoltm-eter (Weston 1240); motor (EC MotamaticModel E600-O13)Impeller: 6-blade MixCo turbine, 4 in diam-eter, 0.8 in above tank, speed 200 .t 2 rpmTank shape: Octahedral; 4.8 in largest dimen-sionSolid: 420.0 g of quartz, density = 2.65

gjcm3Pulp density: 1.6 g/cm3 (67.8% solids byweight)Pulp volume: 0.2 litre

1. Tests to determine the effect of amineon the primary breakage of quartz (singleimpact tests)These tests were conducted in a drop-

weight mill of a similar design to the one usedby other workers [14, 15]. The impact wasapplied by means of a falling mass on a spreadlayer of (10 X 20 mesh) quartz particleswetted by the desired solution. The sizedistribution of the broken products wasdetermined, and the percentage change inamount passing a 20 mesh screen calculated.

RESULTS AND DISCUSSION

Grinding resultsEarlier work [16] has suggested that amine

can influence wet ball milling of quartz bothbeneficially and deterimentally, depending onpH and additive concentration. These resul~are reproduced in Fig. 1. It can be seen thatthe amine addition to the ball mill is benefi-cial, except under acidic conditions, wheredetrimental effec~ are obtained at all aminelevels. Also, in the neutral pH range, additionof amine at levels lower than 10-3mol/l isfound to lead to reduced grinding.

2. Turbidity testsFlocculation characteristics of quartz slimes

were studied by determining turbidity underdifferent chemical conditions. These testswere conducted by measuring the percentageof transmitted light through a suspension. Theinstrument used was a Spectronic 20 spectro-photometer. For this purpose, 0.3 g samplesof -400 mesh quartz were prepared in anagate morter, then mixed with 50 cm3 of10-4 mol/! of ferric nitrate adjusted to thedesired pH, ionic strength, and additiveconcentration to simulate the grindingenvironment. Ferric nitrate was added to

Single impact testsThe effect of pH on the size distribution of

quartz broken by single impact in 10-5 mollldodecylammonium chloride and 10-4 molllFeCI3, at three different pH values, is shown

269

recoveries that are usually obtained for quartzusing amine under similar pH conditions. Theobserved maximum has been attributed in thepast [17 - 19] to the adsorption of theneutral amine molecules along with the ions.Recently, the increase of flotation recoveriesat pH 10.5 has been attributed to the forma-tion of ionomolecular complexes which hassignificantly higher surface activity than thatof the complexing species [20,21]. Similarly,the observed increase in primary breakage ofquartz at pH 10.5 could be related to thepresence of such ionomolecular amine com-plexes in the alkaline pH range.

Torque measurementsThe effect of two levels of dodecylam-

monium chloride on the torque exerted on animpeller in quartz suspension, in 10-4 molllferric chloride solution, at three different pHvalues is investigated. The results indicate adecrease in torque in the amine solution,compared with that in ferric chloride solutionalone, at all pH values. The data obtained atpH 10.5 are given in Fig. 3. It should bementioned that the reduction in torque, dueto amine addition, is more pronounced atpH 10.5 than at pH 3.0 or at pH 5.9. Asdiscussed earlier, this might again be relatedto the formation of highly surface-activeionomolecular amine complexes in thealkaline pH range.

pH

Fig. 1. Effect of amine on amount of -200 meshproduced by wet ball milling of quartz. DDAOconcentration: At 10-5 mol/!; 0, 10-3 mol/l; 6,10-1 mol/!. Ionic strength 3 X 10-2 molll (KNO3).

50:.;

I20

1.1~i 101 ~

/Flocculation studies

Results obtained for the transmission oflight through quartz suspensions in watercontaining 10-4 mol/l Fe(NO3)3 and in aminesolutions of different concentrations are givenin Fig. 4. The data suggest a strong possibilityof quartz flocculation in the presence ofamine even at lower levels of additions. In the10-3 mol/l amine solution, quartz particles arefound to exhibit less flocculation than in10-5 mol/l solution. This is in accordancewith the expected flocculation behavior onthe basis of a higher zeta potential of quartzparticles in 10-3 mol/l than in 10-5 mol/lsolution [22). This is further confirmed bythe theoretical estimates of the stability ratioW from the reported zeta potential values(see Fig. 5). Notice that as W decreases, rapidcoagulation occurs. The values of stabilityratios are calculated using the followingformula [23]:

. . . . . I .. . . . . . . I I~ 10801. 200 -_-,-

~ -. Fig. 2. Gaudin-Schuhmann size distribution of quartz

crushed in (10-5 molll DDACI + 10-4 molll FeO3)solution at different pH values, by a single impact ina drop.weight mill. pH: 0,10.5;6,5.9;.,3.0. Ionicstrength: 3 X 10-2 mol/l (NaCI).

in Fig. 2. The data show that quartz can bebroken more easily at this amine level atpH 10.5 than at neutral (pH 5.9) or acidic pHvalues (pH 3.0). It is interesting to comparethese results with the maximum flotation

~

Fie. 4. Effect of dodecylammonium chloride onthe turbidity or quartz luspenaion in 10-4 mol/lFe(NOJ)J solutiona. Quartz in 10-4 mol/l Fe(NO)hand: e, distilled water; 6, 10-5 mol/l DDACl; 0,10-J mol/l DDACl. Ionic Itreneth 3 x 10-2 mol/l(KNO).

--. -. -Fig. 3. Effect of amine on the torque exerted on animpeller stirring quartz suspension in 10-4 mol/lFeCI, M>lution. Amine concentration: °. 0.0 mol/l;o.10-smol/l;6.10-'mol/l.lnitial pH. 10.5 .t.O.l;final pH. 9.9.t. 0.1. Ionic strength: 3 X 10-2 mol/l(NaCl).

31n(Dl/Ja)2c' +B

4

wherer

12(2ICG + l)Eo

1B = In2..;i- - In

2

where A is the Hammaker constant = 10-12erg for the present system, " is the Debye-Huckel reciprocal length, E is the dielectricconstant, K is the Boltzman constant, T is theabsolute temperature, a is the particle radius,and 1/16 is the stern potential. In our calcula-tions, we have used the zeta potential valuesfor 1/16.

CorrelationsFor breakage to be achieved, particles

should be captured in the grinding zone andsubjected to a fruitful grinding event of eitherimpact or abrasion. In other words, the rateof grinding will depend on the probability ofbreakage, P (breakage), which can be visual-ized as a function of probability of capture,P (capture), and probability of breakage uponcapture, P (breakage/capture), as suggested byHartley et al. [24]. P (breakage/capture) canbe expected to depend on the strengthproperties of the solid while P (capture) canbe expected to depend, among other factors,on pulp fluidity and flocculation.

To foster correlations for the dodecylam-monium chloride/quartz system, the dataobtained under different conditions in thepresence of amine will be compared withthose obtained in its absence in the three pHregions, i.e. acidic (pH 3.0), neutral (pH 5.9)and alkaline pH range (pH 10.5). The changesin different properties of quartz suspensions

211

~,5

2 . 6 8 10

pHFig. 5. Effects of dodecylammonium chloride on thestability ratio W of quartz suspensions in 10-4 mol{lFe(NO3)3 solutions. Quartz in 10-4 mol/l Fe(NO3)3and: e, water; fl, 10-5 mol{l DDACl; 0,10-3 mol{lDDACI. Ionic strength: 3 X 10-2 mol{l (KNO3).

~Fig. 6. Change in different properties of quartzsuspension due to 10-5 mol/l amine addition as afunction of pH. " grindability; °, breakage; D,fluidity;., flocculation.

due to amine addition are shown in Figs. 6and 7.

A. Acidic solutions (pH 3.0)a) The results show a decrease in grinding

rate in the presence of amine at all levels ofamine addition.

b) Amine addition does not affect theprimary breakage of quartz particles in thispH range.

3) Addition of 10-3 mol/l amine does causean improvement in the flow properties of thequartz suspension. Lower levels of amineaddition (10-5 mol/I) show only a smalleffect on fluidity at this pH.

d) At both levels, amine does not appre-ciably affect the flocculation state of quartzsuspension. However, it is possible that theflocs formed in the presence of amine arestronger than those formed in its absence, as aresult of interactions between the hydro-phobic amine chains.

pH

Fig. 7. Change in different properties of quartzsuspensions due to 10-3 mol/l amine addition as afunction of pH. e, grindability; 0, breakage; 0,fluidity; A, flocculation.

Possible mechanismsFrom the above discussion, it appears that

the decrease in grinding rate is due to the

Possible mechanismFrom the above observations, it can be

concluded that the beneficial effect of amineon grinding at pH 10.5 is due to the summa-tion of positive effects on primary breakageand the pulp fluidity. Negative contributionfrom the flocculation is apparently unableto dominate over the beneficial effects.

strength of the floes when amine is used. For-mation of strong flocs could affect grindingowing to:

a) decrease in energy utilization due todisintegration of floes and dislodging of thefines leading to: 1) a decrease in the probabil-ity of coarse particles breakage, P (breakageupon capture), and 2) decrease in the proba-bility of nipping of coarse particles due to theincrease in diameter of particles, P (capture)will, in such a case, be less in the presence ofamine.

CONCLUSIONS

B. Neutral solutions (pH 5.9)a) The grinding data suggest a slight de-

crease in the grinding rate in 10-5 molllamine solution at this pH. Increased addition(10-3 mol/!) enhanced the grinding rate onlyslightly.

b) Primary breakage characteristics remainunaltered by the addition of amine.

c) Amine does not change the pulp fluidityat 10-5 mol/!. However, increased pulpfluidity was observed in 10-3 mol/! aminesolutions.

d) As in acidic solutions, there is no effectof amine on the extent of flocculation of thepulp. However, the flocs themselves can bestronger in the presence of amine due tointeractions between the hydrophobic chainsof the amine species.

Possible mechanismsThe above analysis suggests that the in-

crease in the strength of the flocs in thepresence of amine is possibly the reason forthe detrimental effect on grinding at this pHwhen 10-5 mol/l amine is added. This, alongwith the slight flocculation effects, couldhave also masked some of the beneficialbehavior expected at higher levels (10-3 mol/l)of amine due to the improvement in pulpfluidity at such levels.

Correlation of the grinding results in aminesolutions with the results of basic studiesconducted under simulated chemical condi-tions has led to the following conclusions:

1. The decrease in grinding rate due toamine additions (10-s molll and 10-3 mol/l)at pH 3.0 and pH 5..9 is attributed to theformation of stronger flocs when amine isused (owing to chain-chain interactions).

2. Formation of such flocs could affectgrinding owing to: a) decrease in energyutilization owing to use of it for disintegra-tion of the flocs also; b) cushioning of thecoarser particles.

3. Improved grinding is obtained at pH 5.9.on adding amine at a dosage of 10-3 mol/l.This is attributed to the improvement ob-tained in pulp fluidity due to such amineaddition. The small magnitude of this bene-ficial effect is due to the retarding negativecontribution from pulp flocculation.

4. Marked beneficial effects were obtainedat both levels of amine (10~s mol/! and10-3 mol/!) at pH 10.5. These effects resultedfrom the cumulative effect of improvedprim~ breakage and the flow properties ofthe pulp. Negative contribution from floccula-tion at pH 10.5 was not sufficiently large tomask the beneficial effects.

5. It is suggested that at pH 10.5 amineaffects various interfactial properties signifi-cantly and, consequently, the grinding, owingto the formation of highly surface-activeiono-molecular complexes in this pH range.

ACKNOWLEDGEMENTS

The authors wish to acknowledge thesupport of the American Iron and SteelInstitute.

C. Alkaline solutions (pH 10.5)a) The data at pH 10.5 indicate an increase

in the grinding rate.b) Primary breakage is increased by the

addition of amine in this pH range.c) The pulp fluidity is better in the pres-

ence of amine.d) Quartz particles are flocculated in amine

solutions.

273

REFERENCES 12 R. R. Klimpel and L. G. Austin, Powder Technol.,31 (1982) 239.

13 R. R. Klimpel, Powder Technol., 32 (1982) 267.14 H. E. Rose, 10th Int. Mini. Proc. Congress,

London,1973, pp. 143 - 170.15 A. Z. Frangiskos and H. G. Smith, Tram- Mini.

Dressing Congress, Stockholm, Sweden, 1957, 67.16 H. EI-Bhall, S. Vidange and P. Somasundaran,

Int. J. Min. hoc., 6 (1979) 105.17 F. Aplan and D. W. Fuerstenau, in D. W. Fuerste-

nau (Ed.), Froth Flotation, 50th AnniversaryVolume, AIME, 1962, p. 196.

18 R. W. Smith, Trans. AIME (1963) 226.19 H. Schubert, Freiberg, Forsch., A 355 (1965) 51.20 P. Somasundaran, Int. J. Min. Proc., 3 (1976) 35.21 K. Ananthapadmanabhan, P. Somasundaran and

T. W. Healy, The Chemistry of Oleate and AmineSolutions in Relation to Flotation, paper pre-sented at 107th Annual AIME Meeting, Denver,78-B-68 (1978).

22 H. EI-8halI, A. Gorken and P. Somasundaran,Effect of Chemical Additives on Wet Grinding ofIron Ore Minerals, paper presented at the xmInternational Mineral Processing Congress, War-saw (June 1979).

23 P. Somasundaran, T. W. Healy and D. W. Fuerste-nau,J. Colloid and Interface Sci., 22 (1966) 599.

24 J. N. Hartley, K. A. Prosbreyand O. J. Wick,Energy Reduction in Ore Grinding, paper pre-sented at the 83rd Annual Convention of theNorthwest Mining Association, Spokane, Washing.ton (Dec. 1977).

1 P. A. Rehbinder, Physik, V. 72 (1931) 191.2 A. R. C. Westwood and D. L. Goldheirn, J. Appl.

PhysiC8, 39 (1968) 3401.3 H. E. Rose and R. M. E. Sullivan, Ball, Tube and

Rod Mills, Chern. Pub. Co., New York, 1958,p.30.

4 K. I. Savage, L. G. Austin and S. C. Sun, Trans.AIME, V. 225 (1974) 89.

5 R. Klimpel and W. Manfroy, Deuelopment 01Chemical Grinding Aids and their Effect onSelection-lor-Breakage and Breakage DistributionParameters in the Wet Grinding 01 Ores, paperpresented at Xllth Int. Min. Proc. Cong. SaoPaulo, Brazil, 1977.

6 K. M. Hanna and A. E. Gamal, Powder Technol.,17 (1977) 29.

7 R. R. Klimpel and W. Manfroy, in S. Ramani(Ed.), 14th Int. Sym. on Application. 01 Com-puters in the Mineral Industry, AIME, New York,1977, pp. 197 - 206.

8 R. R. Klimpel and W. Manfroy, Ind. Eng. Chern.Process Des. Dev., 17 (1978) 518.

9 R. R. Klimpel and R. Samuels, Proc. AnnualMeeting 01 Canadian Mineral Processors, Ottawa,1979.

10 R. R. KIimpel, in P. Somasundaran (Ed.), Proc.Sympo,ium on Fine Particle, Procesling, LalVegal, 1979, AIME, New York, 1980, p. 1129.

11 M. Katzer, R. Klimpel and J. Sewel, MiningEngineering, 33(1981) 1471.