Powder TechnololY, 38 (1984) 275 - 293 275

Physico-Chemical Aspects of Grinding: a Review of Use of Additives

H. EL-SHALL

Montana College of Mineral Science and Technology, Butte, MT 59701 (U.S.A.)

and P. SOMASUNDARAN

Henry Krumb School of Mines, Columbia University, New York, NY 10027 (U.s.A.)

(Received August 20, 1982;in revised form August 1,1983)

SUMMARY

The efficiency of energy utilization intumbling mills is discussed in terms of com-ponent processes taking place inside the mill.The past reports on the effect of physico-chemical parameters of the environment onmechanical properties and grindability ofmaterials are reviewed. Reported mechanismsexplaining such effects are analyzed andpossible ways to improve the grinding effi-ciency, through the use of chemical additives,are also discussed.

EFFICIENCY AND ENERGY CONSUMPTION INORE GRINDING

Grinding is an important industrial opera-tion that is used for the size reduction ofmaterials, production of large surface areaand/or liberation of valuable minerals from.their matrices. In addition to mineral proces-sing, it is widely used in the manufacture o:fcement, pigments and paints, ceramics,pharmaceuticals, and cereals. However, theefficiency of this operation is very low [1].

In mineral beneficiation, grinding is alsothe most energy~onsuming process. Energyconsumed in ore grinding in various mills ascompiled by Hartley et al. [2] is shown inTable 1. It is to be noted that the energyconsumed in grinding alone represents up to70% of the energy for the whole beneficiationprocess (see Table 2 reproduced from a 1934article [3]). Corresponding figures for thepresent operations can be expected to bemuch higher. The decrease in ore grade andconcomitant increase in the degree of finenessof values in the ore increase also the need for

r

fine and ultrafine grinding. Consequently, theenergy consumption is now higher, and thegrinding cost represents a relatively higher

..portion of the total beneficiation cost. There-lore, it becomes an important task to improvethe energy utilization inside the grinding mill.

A scheme for the flow of energy and itSutilization in tumbling mills is shown in Fig. 1[ 4 - 6]. It can be seen that grinding energy isexpended for 1) elastic deformation, 2)plastic deformation, 3) lattice rearrangementsand mechano-chemical reactions, and 4) newsurface energy. External to the particles,energy can be expended for 1) frictionbetween the particles and the grinding mediaas well as between particles and particles, 2)~ound energy, 3) kinetic energy of theproducts, and 4) deformation and wear of thegrinding media. The scheme shown in Fig. 1,and the data given in Tables 3 and 4 suggestthat the actual energy needed for fracture(i.e. to produce new surface area) is only asmall fraction (less than 1%) of the totalenergy input to the grinding mill. A greatproportion of the energy input (more than75%) is lost as he~t, probably due to friction,non-productive collisions, elastic and plasticdeformation, etc. However, as considered byPiret [8], the energy expended in elastic andplastic deformations might be necessaryactivation energy for subsequent fractureprocess. If this deformation energy is in-

. cluded, grinding efficiency can be consideredto be 20 - 50% [6].

The above discussion suggests that energyutilization can be improved by using suitablemeans for minimizing the crushing media andthe interparticle and pulp losses. Evidently, aclear understanding of the grinding com-ponent processes and their dependence on the

\

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

TABLElEnergy consumed in grinding at various mills [2J

Concentrate produced Plant En(k'-8.7.78.9

11.7.

10.10.1311107

1768

152095746777

1447

10794

19

..~

CuCuCuCuCuCuCu

Cu-FeCu-FeCu-FeCu-FeCu-NiCu-NiCu-Ni

Pb-Zn, Cu-Zn, ZnPb-Zn. Cu-Zn, ZnPb-Zn, Cu-Zn, ZnPb-Zn,Cu-Zn,ZnPb-Zn, Cu-Zn, ZnPb-Zn, Cu-Zn, Zn

Pb,ZnPb,ZnPb,ZnPb,ZnPb,ZnPb,ZnPb,ZnMoAuAuAuAuAuAu

Calumet & Hecla, MichiganAnaconda Co., MontanaP. D. Lavendar Pit, ArizonaMulfulira, N. RhodesiaSilver BellGalena (Idaho)Copper Cities

Cyprus Mines, CyprusKeretti, FinlandBoliden, SwedenSan Manuel, ArizonaKotalahti. FinlandInco-Copper Cliff, OntarioInco-Creighton. Ontario

Tennessee CopperSt. Joseph Indian CreekA.S. & R BuchansA.S. & R PageIdaradoGilman

St. Joseph Lead. MontanaBroken Hill. AustraliaNew Broken Hill, AustraliaNorth Broken Hill, AustraliaBroken Hill South. AustraliaSoc. Algerienne, MoroccoBunker Hill, Idaho

Climax ColoradoCarlin Gold Mining CompanyBenguet ConsolidatedCortez Gold MinesHomestake Mining CompanyRosacio Dominicana, Dominican Rep.Hogon~Suyoc Mines, Incorporated

,

environment in the mill is needed in order toidentify possibly ways for improving thegrinding efficiency.

GRINDING COMPONENT PROCESSES

of the products to the discharge end of themill. Any parameter affecting any or allof these processes will affect the grindingefficiency.

Breakage of a particle can be achieved ifthe particle is captured in the grinding zoneand subjected to a fruitful breaking action.Probability of breakage. P (overall breakage).

, is thus the product of probabilities for theabove two basic processes [2]:P (overall breakage)

= P (capture) X P (breakage upon capture)The probability of capture. P (capture). is

the probability that a particle will be capturedin a grinding zone and is expected to depend.

\

Two major simultaneous component pro-cesses, namely pulp flow and stress applica-tion, are involved in grinding operations.Specifically, these processes include transportof material to the grinding zone, and subject-ing the material to the grinding actions,leading possibly to propagation and/orinitiation of cracks, followed by the transport

ergyNh/t)

.5

.5

.0

.2

.1

.2

.0

.9

.7

.3

.2

.5

.0

.2

.2

.8

.8

.8

.8

.9

.5

.1

.3

.0

.3

.9

.7

.9

.11

.5

.5

.5

~

277

TABLE 2Energy consumed in grinding in different plantscompared with the total energy needed for the wholebeneficiation [3]

TABLE 3Experimental measurements of crushing and surfaceenergies [6 ]

Material Crushingenergy(erg/cm2)

Surface

energy(erg/cm2)

Power consumedin grinding('Yo plant total)

Plant

\. -Quartz (8iOvGlass (8iOVCalcite (CaCOJ)Halite (NaCl)Barite (8a804)

11700082000324002610071600

92012101100276

1250

305060355645575959

TABLE 4Distribution of energy in a ball mill (7 J. Reproducedby permission from G. C. Lowrison, Crushing andGrinding, Butterworths, London, 1979.

38653646

Energy distribution Energyconsumpti,(%)

A Copper oresHaydenHarmonyUnited VerdeCopper QueenOld DominionArther and MagnaEngeloWalkerBritannia

B Lead and zinc oresChief consolidatedTyboHughesvilleMorning

C Gold oresSpring HillPorcupineKirkland Lake

564470

Bolt frictionGear lossesHeat losses from drumHeat absorbed by air circulationTheoretical energy for size reduction

4.38.06.4

31.00.6

97.6

among other things, on the fluidity and theparticle transport in the mill and the floccula.tiOD of particles. The probability of breakage,

provided that capture has occurred, P (break.age upon capture), is related to the particlestrength. Additives can affect grinding rates

fR>ss ~Y JII'Ur- -t ICTICI~!(,E.~

rRICTICI 1K.E. AlII F-.t.. ,tlASTlC III> PlASTIC:I~F~TICIl~

T-ISSI(lllJ)SSES -I8T

tt:r EtERGY INPUT ro KILL .

It,

.t

losSf;s IN CR\$1ING IEDIANI> OF CRUSHI/; 9SIF~S

IQr

&a;y INPUr TO TIE OMGE~~ -i ICTIOO IHrEII9MTIClE NIl P\U~S K.E. AICI P.E.

: ~I R~TI(JI! - __~TE DI SINTESIAT 100

; IbT

~F~ EI6IGV

CllUATI\t: B8GY IIAIr 10 TIEINOIVlnk PARTIa.ES OF TIE OMGE

fRulTlSS Ir-=TS NC> STA~S

ifRlCTlm t-K.E. NIl P.E.ElASTIC NIl PLASTIC

~TlmfNew;y AESI1.TIIG IN fRACTIR.(Nt) ~~TE DISINTESlATlOO)

ii,t- FRK.TIIIE -

\

tRICTIOI IK.t,* NG> P.E,-J tLAsrlC Nt> PLASTIC ~T

I~TIOI :l~ICAL flfK:rlOlS :~ J

-~ =t10l J ma~S~ACE ~Y

Fig. 1. Energy utilization in model tumbling mill: gross energy >N > C> H> F [4J. (-Residual kinetic energyafter possible secondary fracture has occurred.\

278

through viscosity, flocculation and agglomera-tion (i.e., affectingP (capture.

.,

Pulp flow processThe transport of material inside a mill

depends primarily on the pulp fluidity, whichis influenced, among other things, by the stateof aggregation or dispersion of the fineparticles inside the mill, determined essen-tially by the nature of interactions betweenthe particles and the grinding media and thosebetween particles themselves.

Modification of the flow of pulp in thegrinding mill has been considered to havespecial potential for increasing the efficiencyof the grinding process. This is possiblybecause the fluidity of the pulp can determinehow well particles are transported to regionswhere grinding action is most severe [2]. Inthis regard, the effect of the grinding environ-ment on the pulp fluidity and consequentlyon the grinding efficiency cannot be ignored.

plastic deformation in the vicinity of thecrack [10,11]. Another important parameterthat might affect the strength of materials isthe radius of the crack tip. The sharper thecrack tip, the higher the localized stress'concentration and, consequently, the lowerthe strength of the material. It is important tonote here that the surrounding environmentmay react with the crack surfaces, leadingprobably to changes in the length and/or theradius of the crack and, as a result, in thestrength of the particles.

Clearly, grinding efficiency depends primar-ily on factors that affect these componentprocesses. Unfortunately, the present under-standing of such processes and their inter-actions with the mill environment is limitedand mostly intuitive or speculative [12].Reported effects of the physico-chemicalenvironment on mechanical properties andgrind ability of materials and mechanismsconsidered as responsible for the observedefforts are disucced next.

EFFECT OF THE PHYSICO-CHEMICAL ENVmON.MENT ON MECHANICAL PROPERTIES ANDGRINDABILITY OF MATERIALS

Rehbinder and co-workers [13] wereprobably the first to conduct a systematic Iinvestigation on the effect of liquids with andwithout additives on the failure of solids.They claimed that liquids. in particular water.played an active part in the process of failureand that this effect could be amplified byadding surface active agents. Since then,various researchers have tested the- effect ofdifferent additives on mechanical propertiesof solids and on various comminution pro-cesses, i.e. drilling. grinding, etc.

Effect of environment on mechanical proper-ties of materials

The effect of surface active agents on thedefonnation behavior of a rock under varioustriaxial compression loads has been studied,among others, by Boozer et at. [14]. Theirresults indicated a decrease in the ultimatestrength of sandstone when the pores weresaturated by oleic acid or oleylamine. Thiseffect was attributed by the authors to theadsorption of surfactant molecules on therock grains even though no adsorption test

Stress application processBreaking of solids in grinding operations

results essentially from subjecting the parti-cles to stresses in the grinding zone, causingpossibly several fractures in each particle.

For fracture to occur, cracks must bepresent or initiated and propagated. Allnatural materials have defects as cracks, flaws,or dislocations. These defects will act as stressmodifiers leading to strength values lowerthan the theoretically expected strengths.

Application of a stress on a body activatesthe propagation of cracks creating surface.Griffith [9J has considered the theoreticalenergy balance for this process by equatingthe energy loss from the elastic strain fieldto the gain in surface energy. The minimumstress required for fracture, according toGriffith's law, is given by

a=l/VT

where E is Young's modulus, "Y is the freeenergy of the created surface, and L is thecrack length.

In actuality, energy will also be consumedfor plastic deformation of the material in thesurface and sub-surface regions. Griffith'sformula must, therefore, contain a surfacestress term that includes, in addition to thesurface energy, the energy consumed for

279

9800

\9400

":"s

a59000r!on

~i 8600E0u

..

~

pH

Fig. 2. Effect of organic and inorganic additives onthe compressive strength of sandstone (ref. 15). "10-4 N dodecylammonium chloride; 8, 10-4 N AlCl);6, 10-4 N Na1COJ.

itself was perfomled. It should be noted thatthere was also no control or monitoring ofcritical parameters such as pH and ionicstrength in this work.

In a similar work, Wang [15] examinedthe effect of organic (dodecylammoniumchloride) and inorganic (AlCI) and Na2CO)electrolytes on the compressive strength ofsandstone. Their results showed that theeffect of the additives was pH dependent, asshown in Fig. 2. The investigators attributedthis to reduction in surface energy upon the

adsorption of such additives on the solidsurface.

Other reports [16] have shown a decreasein the compressive strength of some rocks dueto the presence of water (see Table 5). How-ever, no explanation was offered for theseobservations.

Vutukuri [17] reported the tensile strengthof quartzite immersed in 0.06% aluminumchloride solution to be 11.1% lower than thatin water. However, the data (given in Table 6)have a statistical variance of 30%, whichmakes the conclusion questionable. In an-other investigation, Tweeton and co-workers[18] were, however, unable to find any effectof aluminum chloride when used as anadditive at a concentration of 0.05 ppm onthe tensile strength of fused-quartz thread.

There have been many other investigationsof the influence of environment on mechani-cal properties of materials [19 - 29]. In all ofthese investigations, the decrease in thefracture strength of glass-like materialsdue to different environments has beenfound to range from a few per cent to 300%.Hammond and Ravitz [29], for example, intheir study of the effect of saturated vaporsof water, n-propyl alcohol, acetone andbenzene, on the brittle fracture of silica,observed water vapor to cause a 40 - 50%decrease in the fracture strength from itsvalue in vacuum. In general, their data indi-cated the environments with polar moleculesto cause maximum effects.

These authors attributed the observedeffects to stress corrosion cracking [30, 31].This mechanism is related to that of fractureof metals when simultaneously stressed andexposed to certain chemical environments.

\

TABLE 5Effect of water on the compressi ve strength of rocks [16)

Compressive strength (lbf/in2)Type of rock Saturated/dry(%)

Dry Saturated

244389386

199941661422862225501235437204

12241

251'13715425369

298

5044

1258367

1125680

BasaltBasalt with sandstoneDiabaseDolomiteGraniteGniessLimestoneQuartzite

9718067478335820

.3

.0

.6

.0

.7

.3

.3

.2

280

TABLE 6Effect of AlCl] solutions on the tensile strength of quartzite [17]

Concentration('Yo)

Number ofexperiments

Mean tensile strength(lbf/in2)

Standard deviation('Yo)

Percentage changecompared with water

0 (i.e. water)0.020.040.060.10

2030303030

20062008202717831788

3230373034

I--G. 1-1.111.110.9

TABLE 7Specific environments known to cause stress-corrosion cracking [31

Metal Aqueous environment Comments

Carbon steelsMild steel NO - OH-3 . Also (1) anhydrous liquid NHJ. (2)

SbCIJ. HCl. AlClJ in hydrocarbonMedium strengthUltra-high strength

Stainless steels:AusteniticFerritic

HCNH,%O,H'%

a-, Br-, OH-

Martensitic

a-brass Cupric ammonium complex.NH3 or amine atmospherein presence of H2O and O2

Immune above 50% Ni.Susceptible to hydrogen cracking or

blistering. Immune to CI-, OH-,N03-.

Susceptible to hydrogen cracking.Also, when heat treated to highhardness, to H2O and variousaqueous solutions.

Nitrates may reduce to NH4+ causingscc. S02 is reported to cause scc,but probably causes intergranularcorrosion instead.

H2O/3./3 + 'Y brassAluminum alloys:

4 - 20% Zn-AlCommercial. containing Cu or Mg

H2ONaCl solutions. various

aqueous solutionsH2OMagnesium alloys

Titanium alloys:8% AI, 1% Mo, 1% V, Balance Ti Cl-, Br-, 1- Also H2O, CH3OH, CCl4.methylene

chloride, trichlorethylene at roomtemperature; NaCl at >250 c.

N2O..6% AI, 4% V, Balance TiGold alloys:

Cu-AuAg-Au

Immune above 40% Au.FeCI).Aqua Regia, N~OH. HNO)

the surrounding environment [13, 32 - 381.In this regard, Rehbinder [13] classifiedhardness reducers into two groups. The firstgroup contained inorganic salts such as NaCI,NaOH, Na2CO3, MgCl2, CaCl2 and AICIJ, andthe second group contained organic chemicalswith polar molecules. The effect of theadditives in the first group has been found to

\

Uhlig [31] observed cracking of this kind tobe present only when the metal was undertensile stress in a damaging environment.Table 7 lists specific environments observedby Uhlig to cause stress-corrosion cracking ofmetals.

In addition to strength properties, materialhardness has been known to be affected by

281

a sharp decrease in the surface energy of thesolid in contact with the environment. Besidesinteracting with the crack walls, the environ-ment atoms create a force compatible withthe interatomic bond strength, which pro-motes crack propagation.

In addition to mechanical properties,environment has been found to affect sizereduction of materials. The following para-graphs summarize the published reports, inthis regard.

,

increase rapidly with the level of addition,reaching a maximum at a given level for eachreagent. For example, concentrations of0.01% to 0.05% AICl3 were reported byRehbinder [13] to produce maximum effecton quartzite rocks. The influence of organicchemicals are, on the other hand, found todepend mainly on the molecular weight. Forexample, in a homologous series of organicacids, hardness reduction has been found toincrease with increase in molecular weight.Subsequent work by Montrove [34] on theeffect of xylene and benzene on the hardnessof muscovite showed this mineral to be harderin such environmen~ than in air.

Interestingly, even water and moist air havebeen reported by Westbrook and Jorgensen[35, 36] to decrease hardness of differentmaterials such as oxides, silicates, sulfides,fluorides, carbides, and carbonates.

Westwood and co-workers have done acomprehensive study on the effect of dif-ferent environmen~ on the hardness of MgO[37] and CaF2 [38]. They found the hardnessof these two minerals to increase or decrease,depending on the surrounding environment.The authors ascribed the observed effec~ toadsorption-induced changes in the flowproperties of the near-surface region ascontrolled by the movement of surfacedislocations.

In other investigations [39,40], the sameresearchers correlated the observed effec~with the surface charge properties of thematerials in different environmen~. Theyfound MgO and quartz to be harder (i.e. lessdislocation mobility) at their correspondingisoelectric poin~. In contrast, Engelmannet al. [41] used the oscillating pendulumtechnique to measure the hardness of granitesample in AlCl3 and oleylammonium acetatesolutions, and fo~nd that granite was easilypenetrated at the isoelectric point.

In a recent investigation, Shchukin andYushchenko [42] studied environment-sensitive mechanical behavior of materials onan atomic scale using a molecular dynamic(MD) simulation of strain and failure of acrystal. They have undertaken a series ofexperimen~ to reveal the environment-induced qualitative changes in the atomicpicture of strain and failure. The resul~indicated that adsorption-active atoms oftencaused brittle cracking. This was attributed to

Effect of the environment on grindingSize reduction of solids in grinding pro-

cesses is achieved by subjecting particles todifferent stresses, leading ultimately to thefracture of particles.

Fracture may proceed within the particleitself (intragranular fracture) or along thegrain boundaries (intergranular fracture).While intragranular fracture is sufficient forsize reduction processes, it is intergranularfracture that is required for liberation. Gener-ally, fracture process involves rupturing ofchemical bonds to create new surfaces. Anyphenomenon that can help in breaking of thechemical bonds and retard rejoining of theruptured surfaces can be expected to help thegrinding process [1].

Grinding has, in the past, usually beenconsidered as a physical process controlledonly by the mechanical conditions of thegrinding system. Researchers in the field didnot pay as much attention to the effects ofthe physico-chemical parameters of thegrinding environment on the grinding effi-ciency. Only after Rehbinder [13] hadreported his findings on the effect of chemi-cals in enhancing the mechanical failure ofmaterials, did researchers began to studycomminution processes in the presence ofchemical additives.

Lowrison [7] has compiled a list of addi-tives used in grinding investigations togetherwith the changes claimed (see Table 8). Thedata suggest that as much as a 20-fold increasein grinding rate can be obtained by usingchemical additives. It is the purpose of thefollowing discussion to review reportedeffects of environment on grinding processesand mechanisms that are responsible for theseeffects.

282

_TABLE 8Additives used in comminution (7). Reproduced by permission from G. C. Lowrison, Crushing and Grinding,Butterworths, London, 1979.

Material ground Wetor dry

Additive % added Grindingratefactor* t

DDD

Water 0.060.060.04

1.~1.3

Ceramic enamelsMarbleCement clinker

Alcohols and phenolsMethanolIsopentanol

-1.

20.1.

20.o.

:~.

-

0.010.25

D

QuartzQuartzIron powderQuartzSoda lime glassIron powderCement

sec-OctanolSeries normal alcoholsGlycero.lPhenols and polyphenois

Gypsum

Cement clinker D0.2

-0.020.02

KetonesAcetone

AminesTriethanolamineFIotigan (C12-C14 amine ex coconut oil)

-QuartziteLimestone

2...2t.?

SuI phonic acidsArylalkyl suI phonic acid (RDA) 0.06 Graphite

Cement 1.3

Fatty acidsOleic acid 1.10.003 Limestone

Zinc bIen deQuartzPumiceLimestoneQuartzLimestoneLimestone

1.27Butyric acidStearic acid 1

0000

0

00

5

2.02.0

Sodium oleate

Vinsol resin (calcium stearate)

Sodium stearate1.2

Aluminium stearateCaprilic acidMarine oilBeef tallowWool grease

DolomiteCement clinkerCementChrome ore-magnesiteChrome ore-magnesite

DD

1.21.1

LimestoneGypsum

0.11.01.0

Cement clinkerQuartziteQuartzite

DWW

1.331.401.80

DSoda lime glass

1.23Quartz

\ 1.3

Other carboxylic acidsNapthenic acidSodium napthenateSodium sulphonapthenate

Hydrocarbonsn-alkanes, various

EstersAmylacetate

OthersCarbon black 0.08

0.32CementCement

(continued)

290405

.0

.1

.1

.05

.10

.15

.5

.5

.0

283

TABLE 8 (continued)

Additive % added Material ground Wetor dry

Grindingratefactor.

Sodium silicateSodium hydroxide

1.16\--1S1

-1

--

-o.o.o.

o.

Sodium carbonateSodium chlorideCarbon dioxide

wAluminum chloride

-

1.3

Ammonium carbonate

Clay slipMagnesiteLimestoneLimestoneQuartziteMagnesiteDolomiteQuartziteCarbon blackGraphiteTalcMicaVermicullitePumiceLead -zinc ore

-0.01-0.04

10.00.02

Hardwood pitchSodium polymetaphosphate (Calgon)

SulphurQuartzite

DW

-1.65

KaolinThalium chloride-

-Grinding rate factor = new surface produced with additive/new surface produced without additive.

Grinding in waterWet grinding is generally more efficient

than dry grinding r 43 - 46 J. This has beenattributed mainly to chemical reactionsbetween broken surface bonds and watermolecules r 46 J. On the basis of this consi-deration, water vapor should also be expectedto cause similar effects since it can easilypenetrate to the crack tip. Ziegler r 47J hasfound that water vapor can increase theefficiency of cement dry grinding in a smallvibratory mill, but he also observed organicvapor to produce such effects. Additionalevidence has been provided by Locher et at.r 48 J, who have reported that the grindingrate of soda lime glass was higher in humid airthan in vacuum. In contrast to the chemicaleffects discussed above, investigators have alsoconsidered physical factors such as reducedcushioning in water of coarse particles and thegrinding media by the fines as well as effects ofviscosity and specific gravity [1, 12,49 - 76].

\

Grinding in organic liquidsThere have been several reports indicating

the grinding in organic liquids to be moreefficient than in water [46, 51 - 57]. Forexample, Kiesskalt [51] has obtained a12-fold increase in surface area for grindingin organic liquids, such as isoamyl alcohol,

from that in water. In a similar investigation,Engelhardt [53, 54] studied the effect ofalcohol on quartz grinding. His results in-dicated that in the case of alcohol, less energywas needed to produce the same surface areathan that required in water.

In 1956, Ziegler [47] reported a 40%increase in the capacity of cement grindingmills by the use of organic liquids such asethylene glycol, propylene glycol, butyleneglycol and triethanol amine. Since then,organic liquids have been reportedly used ingrinding of cement commercially [52]. It wassuggested [52, 55] that the vapors of theorganic liquids reduced the forces of adhesionbetween the particles and that led to reducedagglomeration and enhanced material flowand, in turn, to improved grinding.

In another investigation, Lin and Metzmager[46] ground quartz in different environments.Their data showed as much as a 25% increasein grinding rate in carbon tetrachloride andmethyl cyclohexane than in nitrogen. Grind-ing in water was more efficient than in anyof the other environments tested. Inter-estingly, when water was added to the organicliquids, the efficiency of grinding was restoredto its value in water alone.

Other organic liquids have been used bySavage et al. [57] to grind silicon carbide in a

.5

.0

.2

.55

0080208

03

284

TABLE 9Effect of surface active agents on grinding of quartzite and limestone [59 ]

Feed Additive Additive concentration(%)

Percentage increasein new surface

Flotigam PQuartzite 00.0050.010.030.050.11.0

00.0080.00170.0030.0060.061.33

+1+1+1

+++'

+

-i

"-',-,'~

'C;

~~

c~++'

,

Limestone Oleic acid

Limestone Sodium oleate 0o.

o.

O.

1.

0

O.

O.

O.

O.

Limestone Flotigam P

laboratory vibratory mill. They obtained a100% increase in the grinding rate in benzenein comparison with that in water alone, whilethe effect of ethanol was considerably less.

Effect of surfactant addition to grindingenvironments

There have been several reports in thepast that grinding efficiency can be influencedby the addition of surface active agents intothe grinding mill. For example, Flotigam PC12-C14) amine from coconut oil) was foundby von Szantho [58] to produce as much as a100% increase in the surface area of quartzand 75% increase in the surface of limestonewhen the reagent concentrations were in therange of 0.005 to 0.02% (see Table 9). How-ever, above the concentration of 0.02% of theadditive, the effect was noticed to be reduced.Similar results were obtained with an oleicacid/limestone system except that in thiscase detrimental effects were observed abovea concentration of 0.003% oleic acid. On theother hand, sodium oleate decreased thegrinding rate of limestone at all levels of

addition (Table 9). It should be mentionedhere that the method used for measuring thesurface area of the ground product was notdescribed in the paper. In addition, it appearsthat there was no proper control of importantvariables such as pH and ionic strength in thiswork. The observed decrease in surface areaabove a certain concentration of the additivescould be due to experimental artifacts intro-duced by possible aggregation of the groundproducts. There is no discussion in the paperon the state of dispersion of the groundproducts or on any microscopic examinationof the products.

In a similar investigation, Gilbert andHughes [59] studied the effect of additions ofArmac T (a mixture of saturated and un-saturated CI6 and CI8 amine acetate) on theball milling of quartz. In this case, the addi-tive was found to produce, at a concentrationof 0.01%, a 4 to 12% decrease in the amountof minus 200 mesh produced in the pH rangeof 2 to 5. An increase in the surfactantconcentration from 0.01% to 0.1% decreasedthe grinding efficiency further at all pH

\

0031423716558

00

+710176199

035-18492

0295674+9

.2

.7

.4

.4

.8

.3

.3

.6

.2

.3

.9

.7

.8

.5

..

.2

.4

.6

.6

040850

005010205

285

\

(a)

surfactant. The importance of proper disper-sion of the ground products has been verifiedrecently by EI-Shall et al. [60]. Using micro-scopic examination of selected size fractions,the authors have found that sodium silicatewas not a sufficiently effective dispersant forthe product ground in 10-1 mol/I aminesolution (see Fig. 3(a. The grinding resultsin this case suggested initially the use ofamine to be detrimental (see Fig. 4). How-ever, after proper dispersion of the groundproducts using alcohol/acetone washes (seeFig. 3(b, amine was observed to have actedas a grinding aid.

Other reports of the effect of surfactantaddition include that of Kukolev et al. [61,62], who obtained improved grinding effi-ciency of alumina on using organo-silicone ata concentration of 0.005%. Malati and hisco-workers [63] and EI-Shall et al. [64]have independently produced finer productsby adding oleic acid to the wet milling ofhematite.

The influence of dodecylammonium chlo-ride on grinding of quartz in a porcelain ball

(b)Fig. 3. Photomicrographs of (-14 +20 mesh) sizefraction of quartz ground in 10-1 molll amine in astainless steel mill [60]; (a), dispersed using 8 g/lsodium silicate solution; (b), dispersed using 8 g/lsodium silicate solution and then washed with alcoholand acetone.

,

values. For example, the level of addition of0.1% at pH 5.0 produced as much as a 36%decrease in the amount of 200 mesh producedas compared with that produced in wateralone. The authors have not offered anyexplanation for the observed effects. Soma-sundaran and Lin [1] have examined thesedata and suggested that physical adsorption ofArmac T (cationic surfactant) on the quartzsurface (negative above pH 2.0) could havedecreased the charge on the particles andthereby caused an increase in their floccula-tion. According to them, it is not clearwhether the reported low grinding efficiencywas due to any experimental artifact intro-duced by such aggregation of the groundproducts, or the direct result of a change ininterfacial properties brought about by the

\

286

ore grinding, due to the possibility of thepresence of extraneous ionic species.

Recently, EI-Shall et al. [64] have ob-tained finer products in the neutral andalkaline pH range, on adding dodecylam-monium chloride during quartz grinding in astainless steel ball mill. Detrimental effectswere obtained in the acidic solutions, asshown in Fig. 5. The authors correlated theirresults with the effect of amine on interfacialproperties of quartz (zeta potential andflocculation characteristics) under simulatedchemical conditions. Ferric salts were addedin this case to simulate the iron released fromthe balls and the mill during grinding. This isvery important since ferric ions can evenreverse the charge on quartz particles andconsequently affect the flocculation proper-ties, as shown in Figs. 6 and 7.

11

mill has been investigated by Ryncarz andLaskowski [65]. The effect of the additivewas found, in this case, to depend on itsconcentration and the pH of the solution. Theauthors correlated their grinding results withthe zeta potential of quartz in amine solu-tions. They concluded that grindability wasminimum under conditions of zero zetapotential. It should be noted that the investi-gators have not taken into account thepossibility of surface contamination byaluminum species from the porcelain mill andconsequent effects on the interfacial proper-ties of the ground material. Aluminum speciesat concentrations as low as 10-5 molll havebeen found to reverse the zeta potential ofquartz fines [66]. Adsorption of cationiccollectors (e.g. amine) can indeed be pre-vented under such conditions. In this regard,Ryncarz and Laskowski [65] have pointedout that grinding was extremely sensitive tothe concentration of iron species present intheir grinding system. Although the authorshave concluded that wet grinding of a givenmineral could be improved by the properchoice of additives, they have also warnedthat it would appear to be almost impossibleto predict any effects accurately in the case of

..+30

+25

+20

1E~'0'C

I.5

fu~

~ ~I10

Effect of addition of inorganic electrolytes togrinding environments

The effect of inorganic electrolytes ongrinding of minerals has been investigated byseveral workers [58, 77, 78]. Although resultsreported in the literature are often contra.dictory to each other, grinding has in generalbeen found to be more efficient in the pres-ence of inorganic electrolytes [19].

In 1948, Kukolev and Melnishenko [79]studied the effect of caustic soda and soap onthe ball milling of magnesite (MgCO3) anddolomite (MgCO3. CaCO3). They found causticsoda to be beneficial to the grinding ofmagnesite, with no effect on the grinding ofdolomite. Soap, on the other hand, was foundto help the grinding of dolomite, with noeffect on magnesite. The authors attributedthe observed effects to the difference inadsorption of sodium hydroxide on the twominerals. Somasundaran and Lin [1] havenoted that there is no reason for sodiumhydroxide not to have the same adsorptioncapacity on both minerals. They supportedtheir argument with the help of the reportedbeneficial effects of the above reagents on thegrinding of limestone, which is chemicallysimilar to dolomite [80,81].

In another work, Frangiskos and Smith[80] investigated the influence of sodiumhydroxide and sodium carbonate on thegrinding of limestone using a laboratory drop-weight mill. These workers obtained improvedgrinding efficiency in the presence of the

I ' lonIc~3x1O-'M(KNOJ) i

-1&1 . . . . . . I0 2 4 . 8 10 12 14

pH

Fig. 5. Effect of amine on the amount of -200 meahproduced by wet ball milling of quartz [64]. DDACIconcentration: 6,10-5 mol/l; 0,10-3 mol/I; 'Y, 10-1mol/I. Ionic strength: 3 X 10-2 mol/I (KNO3)'

+15

+10

+5

0

287

80

+40

\~

+20

'>oS:! 0cu

i

...N-20

-40

-60

2 . . 10pH

Fig. 6. Effect of 10-4 molll FeClJ on zeta potential of quartz as a function of pH [67].", water; e, FeClJ. Ionicstrength: 3 X 10-1. molll (NaCl.).

above additives. Ghosh and co-workers [81]have subsequently repeated some of Fran-giskos and Smith's experiments. Their dataalso indicated increased beneficial effects butwith a maximum at an additive concentrationof 0.02%. Increase of the concentration to0.04% produced less beneficial effects. Theauthors failed to give any explanation for theobserved maximum. The presence of suchmaximum has also been reported by otherworkers [58], who obtained poor grindingon increasing additive concentration abovecertain levels. Additional evidence is providedby Halasyamani and co-workers [82, 83] intheir investigation of the influence of pH onthe grinding rate of quartz and calcite in asteel ball mill using HCI and NaCH. Maximumgrinding rate of quartz was obtained aroundpH 7.0. In this case, the authors explained theobserved results as due to increased floccula-tion of quartz in the neutral pH range com-pared with the acidic or the alkaline pH range.They claimed that the flocculated material inthe mill was in a better position to receiveimpacts, leading to improved grinding. Itshould be noted that these investigators didnot verify whether the quartz was flocculated.

,

\

Actually, it is not clear why quartz shouldflocculate at any pH other than close to itspoint of zero charge, i.e. pH -2.0 [84]. Inanother work, Mallikarjunan et at. [83]reported an increase in surface area of calcitewith decreasing pH in the pH range of 8 to 4,whereas the surface area of quartz was foundto decrease with pH in the same pH range.This observation could, however, have beendue to better dispersion of the particles sincethe direction of pH change for surface areaincrease was away from the point of zerocharge for both quartz and calcite (2.0 and10.5, respectively) [84,85].

Savage et at. [57] studied the effect ofsodium chloride and various salts of multi-valent ions on the grinding of silicon carbide,using water or ethanol as the grinding fluid.They classified the sal~ into flocculants(MgCI2, CaCI2, and FeCl2, and FeCI3) anddispersants (NaCI and N~207)' The authorsconcluded that in both water and ethanol,dispersants significantly improved the grind-ing rate.

It has been reported by various workersthat multivalent ions do affect the grindingefficiency of minerals. For example, Fran-

288

In addition to the works discussed above,higher efficiency of grinding has been re-ported in mineral processing plants using seawater to make up the pulp instead of theordinary tap water [86 - 88].

n

~!E~~

..

..

2 4 8 8 10 12pH

Fig, 7, Effect of 10-4 mol{l FeCI) on the turbidity ofquartz slimes as a function of pH [67). e, 10-4 mol IIFeCI) blank test; " water; " FeCI), Ionic strength3 X 10-2 mol{l (NaCl).

giskos and Smith [80] have reported anincrease of as much as 50% and 15% insurface area of ground quartz due to additionof 2.0 mol/! AlCl] and CUS04. respectively.Furthermore. they found iron-stained quartzcrystals to grind faster than the clear ones. Onthis basis, they have suggested that ferric ionsmight have affected the fracture of quartzduring grinding. It should be noted that therewas no indication as to whether these investi-gators have controlled the pH or the ionicstrength of the grinding solution. Thus, noconclusion can actually be drawn about themechanisms involved in such systems.

Recently, Hartley et al. [2] have studiedthe effect of sodium hydroxide on wet ballmilling of taconite ore. They obtained anincrease in surface area when the feed was ofnarrow size range; however, when they used afeed of complete size distribution. no effectwas obtained. They assumed the ineffec-tiveness of the additives in the latter case tobe the result of changes in viscosity due to thepresence of fines in the feed.

Effect of the physical properties of thegrinding environment

As mentioned earlier, transport of thematerial through the grinding mill is one ofthe important component processes in thegrinding operation. The material transport inthe mill depends primarily on physical proper-ties of the pulp, such as pulp fluidity, state ofdispersion or flocculation of the pulp, densityof the solids and density of the medium. Inother words, these properties determine howwell particles are transported to the grindingzone itself. In addition, such properties can beexpected to have an effect on the hydrody-namic behavior of particles as well as on thegrinding media such as balls or rods, and con-sequently, on the grinding performance [1].

It has been reported [57, 89 - 92] that pulpviscosity can influence the grinding of ores.Schweyer [89] found grinding in pebble millsto be dependent on the viscosity of themedium up to a certain grinding time, andthen to be independent of it. In a similarwork, Hockings et al. [90] studied the effectof pulp viscosity (using com syrup to controlthe viscosity) on the rate of production ofquartzite fines in a rod mill and found theincrease in the pulp viscosity from 1 to1000 cP to cause as much as a 50% decreasein the grinding rate when the grinding wasconducted for 1 min. The observed effect wasof a smaller magnitude for 2, 3, and 4 mingrinding (see Fig. 8). Additional results in thisarea have been reported by Savage et al. [57],who ground silicon carbide in liquids ofdifferent viscosities and the results indicatedfaster grinding rates in liquids of lowerviscosities.

Fluidity modifiers and dispersants havebeen reported to produce both an increaseand a decrease in grinding efficiency [2,67,68 - 75,92]. For example Hartley, et al. [2]have obtained a marked decrease in grindingefficiency of molybdenum ore when theyadded sodium tripolyphosphate to theirgrinding circuit. On the other hand, Hannaand Gamal [92] found that sodium silicateenhanced grinding of limestone.

289

1.00:

~=:~ ~ ~8 0.3010.20~ 0.80GO

~ ~

~f 0.602'~I -~ i~ ~ 040.-;tE ~~-u 0.20

~.5E

.,;

&

j

o~0.10

0.07'0S o.~~

~ 0.2 mgf,-/ XFS-4272

No~-

0.01~ 50 200 500 1000 2000

Size I_I

Fig. 9. Back-calculated S curves for taconite oreslurries ground in 8 in batch ball mill showing effectof grinding additives [72 J. All slurry densities at84 wt.%. +, experimental (one-size fraction method);-, computer back-calculation.

0.6 mg/g XFS-4272

~ 0.03I'" 0.02u . .. ... ., 5 10 50 100 SOO 1000 5000

VilCooity u, CP.

Fig. 8. Cumulative percentage finer than the startingsize va; viscosity for rod mill grinding of 10 X 14 meshquartzite in corn syrup-water mixtures for variousgrinding times; after Hockings et al. [90 J. " 2405;0,180 s;6, 1205;0,60 s.

~

J

i

~ i':~~,,9 t"v"'" ,

,c

C;",c:i;,~,~!'1'7V

t~.!~_.! ~ ~ +4 & I +2 - ---

~~JT -

i' Steady state (no aidl'2' Polymer added, Poly , shut off~ Polymer addedt FI8d ,... increase~ +10

5 +&i-5

~. ...~E ~.~ 1~-OJ

~I.10" . . , . . , , . , , ,

7 8 9 10 11 N- 1 2 3 4 5 6l' '2' l' -a"t

Time

Fig. 10. Typical response curve for industrial-scalemill {72].

..

Unfortunately, most of the past work isnot of much practical use due to lack ofproper control of relevant variables such assolution, pH, ionic strength, etc., and possiblyowing to experimental artifacts that can ariseout of agglomeration of products that werenot properly dispersed before size analysisafter contacting with surfactants.

An additional experimental problem can bethe frothing of the pulp in the presence ofsurfactants during grinding and the conse-quent change in the hydrodynamic character-istics of the grinding environment, as well aspossible entrapment of the mineral particlesin the froth (see Fig. I!) [60]. The entrappedparticles can remain levitated and escape thegrinding actions, with resultant reduction ingrinding efficiency. This possibility has notbeen taken into consideration by the pastworkers.

In 1970, Klimpel and co-workers [68 - 75]have initiated a research program aiming atidentifying possible grinding aid chemicals.They have conducted several laboratory andindustrial scale tests using fluidity modifiers.Examples of the results obtained in thesetests are shown in Figs. 9 and 10. The datagiven in Fig. 9 suggest an increase in thespecific rate of breakage due to the grindingaid addition. Similarly, the industrial datashown in Fig. 10 indicate that a finer productcould be obtained with approximately con-stant feed rate whenever the aid is added.Also, a constant grinding size is achieved withhigher feed rates, due to addition of thegrinding aid.

The behavior of these reagents has beencharacterized by measuring their effectson the rheology (viscosity) of the groundproducts. Correlation of rheological data andlaboratory grinding results [74,75] has madeit possible for the investigators to identifyslurry conditions when chemical additiveswould increase rates of breakage. In thisregard, they have concluded that chemicalsthat can work as grinding aids should main-tain pseudoplastic behavior in the slurrywithout yield stress, or reduce the yield stressin a dense pseudoplastic slurry.

On the basis of the above discussions, onecan conclude that comminution processescan, in general, be enhanced by the use ofinorganic and organic additives under selectedconditions. Indeed, in order to fully developand utilize such effects, it is necessary toestablish the mechanisms involved in sucheffects.

\

290

11

Fig. 11. Photograph showing frothing during millingin 10-1 molll amine at 78% critical speed, elapsedtime oC grinding = 58 [60).

Unless precautions are taken to avoid or toaccount for such experimental artifacts andproblems, observations, even though real, canbe misleading. In addition, without a fullunderstanding of all the interfacial propertiesof the systems under study, comminutionresults are likely to be attributed directly tothe effect of additives on fracture.

The above review of the past results clearlyshows that most of the past works have nottaken into account the above-mentionedproblems, and therefore no definite mecha-nisms could be derived on the basis of suchresults.

Reported mechanismsMost of the past results of the effect of

chemical additives on comminution of~aterials have been interpreted in terms oftwo major mechanisms. The first is called'Rehbinder's effect' [93J. It is based on thereduction of surface free energy of solidsinduced by the adsorption of surface activeagents. Since fracture of materials in com-minution processes involves creation of newsurfaces, the amount of energy requiredshould be proportional to the surface freeenergy of the created surfaces. If the surfacefree energy could somehow be reduced, thenthe energy consumption in creating the samesurface area could be expected to be less. Onthis basis, adsorption of surface active agentsduring grinding should be expected to im-prove the grinding efficiency. However, asdiscussed earlier, the actual amount of energyneeded to create new surface represents onlyabout 1% of the energy input to the commi-nution system. In addition, the effect of

,.

chemical additives on other important param-eters and properties of the system such asplastic deformation, crack initiation and/orpropagation, flocculation and dispersion, etc.,are neglected in this mechanism.

Primarily, presence or initiation of cracks isthe prerequisite factor for fracture to occur.Additionally, parameters such as crack length,crack tip radius, plastic flow at the crack tip,propagation of the cracks, etc., will determinethe strength of materials. The effect ofchemical additives on such parameters, inaddition to their effect on the surface freeenergy, should therefore be noted.

The second mechanism, based on adsorp-tion-induced mobility of near surface disloca-tions, was proposed by Westwood et al.[37 - 40], who studied the effect of variousadditives on the hardness of different crystal-line and non-crystalline materials. Theycorrelated their hardness measurements anddrilling rates with near-surface dislocationmobility and interpreted 'Rehbinder's effect'as a result of changes in the electronic statesnear the surface, and point and line defectscaused by the adsorption of the additive onthe surface of the solid. Such changes caninfluence specific interactions between dis-locations and point defects which control thedislocation mobility and hence the hardness.However, this mechanism cannot be usedexclusively to explain the effect of theenvironment on grinding, because importantproperties such as pulp fluidity, flocculation,and dispersion, etc., are not considered.

Chemisorption or formation of activatedcomplexes between the additive moleculesand the crack surface has been considered byseveral workers, including Somasundaran andLin [1], to be important to the fractureprocess.

Other investigators [31] have consideredenhanced corrosion due to stress application(stress-corrosion cracking) as a mechanismresponsible especially for the fracture ofmetals. The mechanism responsible forstress-corrosion cracking is not, however,clear at present, particularly because of theinconsistencies in reported results.

Another mechanism considered by someinvestigators [44,57,67,68 - 75, 92] is basedmainly on the role of reagents in dispersionand, hence, the flow of particles in the millduring comminution processes.

291

REFERENCES

.

As discussed earlier, grinding is an integralprocess composed of several si~ultaneoussub-processes, and chemical additives canaffect these sub-processes due to one or allof the previously discussed mechanisms.Identification of the role of each mechanismis possible only if the relevant properties ofthe solid and of the solution in contact withthe solid during its fracture are simulta-neously determined. The experimentallyaccessible properties, in this regard, includezeta potential, surface tension, pH, ionicstrength, temperature, chemical compositionof the solution, and frothing and flocculationcharacteristics of the system. Also, it isnecessary to control and monitor pretreat-ment of the material to be ground and itsphysical as well as structural and surfaceproperties. Finally, the machine character-istics such as speed, size, and mechanisms ofstress application and other related param-eters like loading, material filling, etc., mustbe taken into account.

It is clear from the above review thatchemical additives can influence grinding fora number of reasons:a) modification of the flow of pulp in thegrinding mill;b) influence on reagglomeration of thefreshly produced fines;c) modification of frothing characteristics ofthe pulp during grinding;d) influence of interactions between mineralparticles and balls and mill wall and amongparticles themselves (due to changes infrictional characteristics);e) influence on the strength of the materialdue to: 1) effect on crack initiation andcrack extension energy, and 2) retardation ofrejoining or sealing of the freshly createdcracks.

Elucidation of the mechanisms governingthe effect of environment on grinding de-pends essentially on identification of theeffects on all the above factors and inter-actions between them.

ACKNOWLEDGEMENT

The support of the American Iron andSteel Institute is gratefully acknowledged.

1 P. Somasundaran and I. J. Lin,I and EC ProcessesDes. Dev., 11 (1972)321.

2 J. N. Hartley, K. A. Prisbrey and O. J. Wick,E/MJ (Oct. 1978) 105.

3 A. M. Gow, M. Guggenheim and W. H. Coghill,Review of Fine Grinding in Ore Concentrators,Bu. Mines, I.C. 6757 (Jan. 1934).

4 C. C. Harris, Trans. Inst. Mining and Metallurgy,75 (1966) C50.

5 H. Rumpf, Powder Technol., 7 (1973) 145.6 C. Ou, Jr., in Particulate Technology, MacMillan,

New York, 1966, pp. 44 .86.7 G. C. Lowrison, Crushing and Grinding, CRC

Press Inc., Cleveland, OH, 1979, p. 60.8 E. L. Piret, Chern. Eng. Progress, 49 (1953) 56.9 A. A. Griffith, Philos. Trans. Roy. Soc. London,

Ser. A 221 (1920 -1921) 163.10 G. R. Irwin, Fracture of Metals, ASM Symposium,

Chicago,-1947, p.147.11 E. Orowan, Rept. Progr. Phys., Vol. 12 (1949)

185.12 P. Somasundaran in Jr. G. Onoda and L. Hench

(Eds.), Ceramic Processing Before Firing, JohnWiley, 1978, pp. 105.123.

13 P. A. Rehbinder, Physik, 72 (1931) 191.14 G. D. Boozer, K. H. Hiller and S. Serdengecti, in

C. Fairhurst (Ed.), Rock Mel:hanics, PergamonPress, New York, 1963, pp. 579 .625.

15 F. D. Wang, Ph.D. 1;'hesis, University of Dlinois(1966).

16 M. D. Ruiz, Proc. 1st Congo Int. Soc. Rock, 1(1966) 115.

17 V. S. Vutukuri, Tram. AIME, V. 252 (1972)407.

18 R. T. Daryl, R. D. Tweeton, W. H. Engelmann,G. A. Savanick and D. I. Johnson, U.S.B.M., Rep.of Invest, No. 8186 (1976).

19 A. Joffe, M. W. Kirpitschewa and M. A. Lewitsky,Z. Physik, 22 (1924) 286.

20 V. P. Berdennikov, Physik, Z. Sowjetunion, 4(1933)397.

21 S. J. Gregg, Surface Chemistry of Solids. Rein.hold, New York, 1951, p. 297, Ceram. Abstr.(Sept. 1952) 173e.

22 C. Benedicks and G. Ruben, Jemkontorets Ann.,129 (1945) 37, Ceram. Abstr. (Feb. 1946) 32.

23 T. C. Baker and F. W. Preston, J. Appl. Phys., 1 7(1946)162.

24 N. J. Petch, Phil. Mag. (8), 1 (4) (1956) 331.25 V. K. Moorthy and F. V. Tooley, J. Am. Ceram.

Soc., 39 (1956) 215.26 T. L. Johnston and E. R. Parker, Sixteenth ONR

Tech. Rept. Ser. 127, Issue No. 16, Universityof California, Mineral Research Laboratory(Jan. 1957).

27 W. C. Levengood and W. H. Johnston, J. Chern.Phys., 26 (1957) 1184, Ceram. Abstr. (Oct. 1957)2311.

28 O. D. Sherby, ONR Tech. Rept., ONRL (1958)34.

29 M. L. Hammond and S. F. Ravitz, J. Am. Ceram.Soc., V. 46 (1963) 329.

292

1

59 L. A. Gilbert and T. H. Hughes, SymposiumZerkleinern I, verlag Chern. Duseldorf, Germany,1962, pp. 170 - 193.

60 H. El-Shall, S. Vidarage and P. Somasundaran,Int. J. Mini. Proc., 6 (1979) 105.

61 G. V. Kukolev, I. Ya. Piven, R. G. Zaika, D. V.Minkov, L. V. Vinogradova and T. S. Makarova,Refractories (USSR), 33 (1968) 465.

62 G. V. Kukolev, I. Ya. Piven, V. M. Martynova,D. B. Minkov, A. S. Yudina, T. V. Malikova andB. V. Parkhaev, Refractories (USSR), 33 (1968)741.

63 M. A. Malati, M. K. Orphy and A. A. Yousef,Powder Technol., 2 (1968) 21.

64 H. El-Shall, A. Gorken and P. Somasundaran,Effect of Chemical Additives on Wet Grinding ofIron Ore Minerals, paper presented at the XIIIInt. MinI. Proc. Cong., Warsaw (June 1979).

65 A. Ryncarz and J. Laskowiski, Powder Technol.,18 (1977) 179.

66 M. C. Fuerstenau, D. A. Elgillani and J. D. Miller,Tran,. AIME, 247 (March 1970) 11.

67 H. El-Shall, D.E.Sc. Thesis, Columbia University,New York (1980).

68 R. Klimpel and W. Manfroy, Development ofChemical Grinding Aids and their Effect onSelection-for-Breakage and Breakage DistributionParameters in the Wet-Grinding of Ores, paperpresented at XU Int. Min. Proc. Cong., Sao Paulo,Brazil (1977).

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

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

71 R. R. Klimpel and R. Samuels, Proc. AnnualMeeting of Canadian Mineral Processors, Ottawa,1979.

72 R. R. Klimpel, in P. Somasundaran (Ed.), Proc.Symp. on Fine Particles Processing, Las Vegas,1979, AIME, New York, 1980, p. 1129.

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

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

75 R. R. Klimpel,Powder Technol., 32 (1982) 267.76 U.S. Patent 4,126, 276;4, 126, 277;4, 126,

278;9,136,830;4,162,059;4,162,045.77 J. H. Brown, M.I.T. Progress Report, N.Y.O.,

7172 (1955).78 M. H. Stanzyk and I. L. Feld, U.S. Bureau of

Mines Report 7168 (1968).79 G. V. Kukolevand L. G. Melnishenko, Fireproof

Mater. (Moscow), 13 (1948) 447.80 A. Z. Frangiskos and H. G. Smith, Trans. Miner.

Dressing Congr., Stockholm, Sweden (1957),pp. 67 .84.

81 S. K. Ghosh, C. C. Harris and A. Jowett, Nature,188 (1960) 1182.

82 R. Mallikarjunan, K. M. Pai and P. Halasyamani,Trans. Indian In,t. Metal, 18 (1965) 79.

83 P. Halasyamani, S. Venkatachalam and R. Malli-karjunan, Ind. Eng. Chem. Process Des. Develop.,7 (1968) 79.

30 W. A. Weyl, J. Am. Gram. Soc., V.. 32 (1949) 367.31 H. H. Uhlig, in H. Liebowitz (Ed.), Fracture: An

Advanced Treatise, Academic Press, New York,1971, pp. 646 - 675.

32 V. I. Lichtman, P. A. Rehbinder and G. V.Karpenko, Effects of a Surface Active Medium onthe Deformation of Metals, H.M. StationaryOffice, London, England, 1958.

33 V. I. Lichtman, E. D. Shchuken and P. A. Reh-binder, Physico-Chemical Mechanics of Metals,Academy of Sciences, Moscow, USSR, 1962.

34 M. I. Montrov, Zh. Eksp. Tear. Fiz., 4 (1934) .1957.

35 J. H. Westbrook and P. J. Jorgensen, Trans.AIME, 233 (1965)425.

36 J. H. Westbrook and P. J. Jorgensen, Am. Miner-alogist, 53 (1968) 1899-

37 A. R. C. Westwood, D. L. Goldheim and R. G.Lye, Phil. Mag., 16 (1967) 505.

38 A. R. C. Westwood and D. L. Goldheim, J. Appl.Phys., 39 (1968) 3401.

39 N. H. MacMillan, R. D. Huntington and A. R. C.Westwood, 7th Rept. to Office of Naval Research,Contract No. 0014-70-G-0300 (July 1973) 15.

40 A. R. C. Westwood, N. H. MacMillan and R. S.Kalyoncu, Trans. A/ME, V. 256 (1974) 106.

41 W- H. Engelmann, O. Terichow and A. A. Selim,Zeta Potential and Pendulum Sclerometer Studiesof Granite in a Solution Environment, Bu. MinesRI 7048,1967.

42 E. D. Shchukin and V. S. Yushchenko, J- MaterialsSc., 16 (1981) 313.

43 F. C. Bond, Mining Congress J. (January 1975)38-40.

44 H. E. Rose and R. M. E. Sullivan, Ball, Tube, andRod Mills, Chem. Pub. Co., New York, 1958,p.30.

45 P. Somasundaran and D. W. Fuerstenau, Trans.A/ME, 226 (1963) 132.

46 I. J. Lin and A. Metzmager, Trans. AIME, 241(1968)412.

47 E. Ziegler, Schririenreihe der Zementindustrie,No. 19, Verein Deutscher Zementwerke, Dussel-dorf, Germany, 1956.

48 F. W. Locher, W. Eichartz, H. M. von Seebachand S. Sprung, Extended Abstracts, 163rd Na-tional Meeting of the American Chemical Society,Boston, April 1972, p. 8.

49 P. C. Kapur, A. L. Muler and D. W. Fuerstenau,Can. J. Chem. Eng., 43 (1965) 119.

50 H. J. Oberson and J. H. Brown, A/ME Trans., 220(1961)110.

51 S. Z. Kiesskalt, Ver. Deut. /ng., 91 (1949) 313.52 F. W. Locher and H. M. V. Seebach, I and EC

Process Des. Dev., 11 (1972) 190.53 W. V. Engelhardt, Nachr. Akad. Wiss. Gottingen,

Math. Phys. KI., No.2 (1942).54 W. V. Engelhardt, Naturwissenschaften, 33

(1946)195.55 W. Fischer, Zem. Kalk Gops, 20 (1967) 138.56 H. Schneider, Zem. Kalk Gops, 22 (1969) 193.57 K. I. Savage, L. G. Austin and S. C. Sun, Trans.

AIME, 225 (March 1974) 89.58 E. von Szantho, Z. fur Erzberglass und Metall-

huttenwesen, 2 (1949) 353.

293

90 W. A. Hockings, M. E. Volin and A. L. Mular,Trans. AIME. 232 (1965) 59.

91 B. Clarke and J. A. Kitchner, British Chern. Eng.,13 (July 1968) 991.

92 K. M. Hanna and A. E. Gamal, Powder Technol..17 (1977) 19.

93 P. A. Rehbinder, L. A. Schreiner and K. F.Zhiyach, Hardness Reducers in Drilling - APhysico-Chemical Method of Facilitating theMechanical Destruction of Rocks During Drilling.Akad Nauk, Moscow, USSR (1974), Trans. byCouncil for Scientific and Industrial Research(CSIR), Melbourne. Australia. 1948, p. 163.

~

84 P. Somasundaran, in G. Goldfinger (Ed.), CleanSurfaces: Their Preparation and Characterizationfor Interfacial Studies, Marcell Dekker, NewYork, 1970, p. 285.

85 P. Somasundaran and G. E. Agar, J. ColloidInterface Sci., 24 (1967) 433.

86 M. Rey and P. RaCfinot, Mining World, 19 (1966)18.

87 M. K. Orphy, A. A. YouseC and H. S. Hanna,Min. Mineral Eng., 4 (October 1968) 48.

88 I. J. Lin, J. Au. Eng. Arch. (Israel), 27 (1969) 27.89 H. E. Schweyer, Ind. Eng. Chem., 34 (1942)

1060.

..

"',

\