j m a t e r r e s t e c h n o l . 2 0 1 3;2(4):386–391

Available online at www.sciencedirect.com

www. jmrt .com.br

Review Article

Influence of dispersants on the rheological and colloidalproperties of iron ore ultrafine particles and their effect onthe pelletizing process—A review

Sandra Lúcia de Moraesa,b,∗, José Renato Baptista de Limaa, João Batista Ferreira Netob

a Department of Mining Engineering, Universidade de São Paulo, São Paulo, SP, Brazilb Instituto de Pesquisas Tecnológicas do Estado de São Paulo, São Paulo, SP, Brazil

a r t i c l e i n f o

Article history:

Received 15 March 2013

Accepted 16 April 2013

Available online 7 November 2013

Keywords:

Iron ore

Pelletizing

Binder

Dispersant

a b s t r a c t

The use of binders for iron ore pelletizing is required to: (a) increase the pellet strength

before heating – green strength; (b) prevent the collapse of the pellets during firing, when

the gases generated by water vaporization could cause cracks. Bentonite is the main binder

used in industry, and its binding mechanism in iron ore pellets has been widely studied

and understood. Efforts to solve the problems of using bentonite in iron ore pelletizing have

focused on the employment of organic binders the composition of which presents dispersant

properties. This review critically evaluates the current understanding of the influence of

dispersants on the quality of iron ore pellets, by means of their action on the colloidal

and rheological properties of iron ore ultrafine particles. Emphasis is given to obtaining

parameters for practical application of this action in the pelletizing process in bench and

pilot scales.

© 2013 Brazilian Metallurgical, Materials and Mining Association. Published by Elsevier

and more attractive as the need to obtain ultrafine materials

1. Introduction

The use of ultrafine powders has increased substantially invarious industrial fields, such as minerals, ceramic materi-als, chemical products, pigments and pharmaceutics, amongothers [1].

As a rule, a comminution process (grinding) takes place soas to obtain such materials. The work by Wang and Forssberger[2] gives an international overview on comminution technol-ogy to produce mineral powders, whose particle size, in most

cases, is P80 < 20 �m.

Much of the mentioned technology employs wet processingand makes use of an additive in order to enhance the

∗ Corresponding author.E-mail addresses: sandralm@ipt.br, sldemora@gmail.com (S.L. de M

2238-7854/$ – see front matter © 2013 Brazilian Metallurgical, Materials and Mhttp://dx.doi.org/10.1016/j.jmrt.2013.04.003

Editora Ltda. All rights reserved.

efficiency of the process, either to facilitate the grinding itselfor to reduce wear rates.

Possa [3] indicates that the processing of fine particlesbecame a challenge in the mineral processing since thedecrease of particle sizes is followed by a decrease of theirmechanical forces. This makes the forces related to the elec-trostatic phenomena and to the discontinuity of the medium(viscosity) more significant.

The rheological properties in the mineral processing havebeen studied for many years now, and are becoming more

oraes).

increases [1,3–5].The mining industry makes use of a wide range of chemical

products at various process stages, from inorganic products to

ining Association. Published by Elsevier Editora Ltda. All rights reserved.

o l . 2

sc

ida

abpb

Clib

bLptvwot

sptb

peoiowt

pidtoep

2

2

Rt

cmr

�

((

(

((

(

j m a t e r r e s t e c h n

pecial synthetic polymers, in order to improve the rheologyontrol of the production process.

A review by Pearse [6] summarizes various products usedn mineral processing and their function at each step, i.e.,epressants, collectors, pH controllers, clarifiers, dispersants,nti-caking, binders, and others.

In the case of iron ore mining, due to the recent short-ge of high grade ores, the particle size of the final product isecoming finer and finer. Thus, the iron ore concentrate needselletizing to allow its use in the reduction processes, both inlast furnaces or direct reduction reactors.

Various types of organic binders (carboxymethyl cellulose,MC, guar gum, hemicellulose, starch acrylate, polyacry-

amides, among others) have been studied and implementedn the pelletizing step in an attempt to replace bentonite, theinder traditionally used.

As noted by Moraes [7], Lima [8] studied the use of CMC as ainder in the pelletizing of iron ore. In previous experiments,ima [8] obtained pellets bonded by organic reagents. Suchellets did not develop sufficient strength, particularly afterhermal treatment. Moreover, their production cost would beery high, or their production would demand a large amount ofater to be added so that the pellets would deform under their

wn weight. These problems were solved by adding sodiumripolyphosphate (TPP) together with CMC.

Motivated by the results obtained by Lima [8] Cas-ola and Chaves [9] conducted their research on iron oreelletizing following two approaches: the first one using ben-onite + CMC + TPP; the second one aimed the replacement ofentonite, CMC, and CMC + TPP with alternative products.

The tests showed that TPP alone was more effective in theelletizing process than the mixture CMC + TPP. They hypoth-sized that the pellets develop strong mechanical propertieswing to the binder action on the finer particles. Binder wouldncrease the viscosity, leading to an increase in the strengthf green pellets; and during heating, its predominant effectas on the dispersion of fine ore, with further crystalliza-

ion/bonding of the dissolved grains.Aiming to verify the action of the colloidisant agent on the

elletization process of iron ore concentrate, this review crit-cally evaluates the current understanding of the influence ofispersants on the strength mechanism developed by pelletshrough its action on the colloidal and rheological propertiesf iron ore fines. Special emphasis is given to obtaining param-ters for practical application of this action in the pelletizingrocess in bench and pilot scales.

. Literature review

.1. Rheological parameters

heology is defined as the science that studies the deforma-ion and flow of materials under the action of a force [1,3].

In the case of fluids, viscosity � is a useful parameter andan be defined as the property that a fluid has to resist defor-ation, given by the ratio between shear stress � and shear

ate �, as shown in Eq. (1).

= ��−1 (1)

0 1 3;2(4):386–391 387

Traditionally, fluids can be classified into Newtonian ornon-Newtonian. In the case of Newtonian fluids, the viscosityis constant, and this is variable in the case of non-Newtonianfluids, which means the shear stress varies non-linearly withthe shear rate.

The rheology of suspensions, in the scope of mineralprocessing, is widely discussed [1,3,10–15].

Mewis and Spaull [16], in their review of rheology of con-centrated suspensions, demonstrated the complexity of therheological behavior of slurries, since it includes viscous, plas-tic and elastic phenomena of suspensions, and its importancefor understanding the process data.

A typical mineral processing flowchart involves operationsof comminution, separation by size, concentration, andtransport which are carried out, when possible, in aqueousmedium [17].

He et al. [1] brings a review of rheology of slurries at thestage of wet ultrafine grinding in the mineral industry. In thispaper, the authors point out that the rheology in this step ishighly complex and is influenced by many parameters suchas solid concentration, size distribution and particle shape,temperature and pH of the slurry and the use of dispersants.

The study carried out by Shi and Napier-Munn [13] indi-cates that the effects of slurry rheology affect the GrindingIndex of materials thus affecting their behavior.

Kawatra and Eisele [14] have highlighted that the rheologycontrol of slurry in a grinding circuit requires the control ofthe percentage of solids, particle size distribution and temper-ature, and an on-line measurement of viscosity of the slurry.

Especially in the fine grinding and flotation operations, rhe-ological aspects are widely discussed, but they need to bebetter understood [12].

2.2. Influence of chemical additives

Ultrafine grinding has increased as the demand for ultrafinematerials for industries has. The processing of very fine parti-cles is a problem mainly because of their small mass and highsurface area.

Sivamohan [18] reviewed the difficulties in recovering veryfine particles in mineral processing. In this review, the authorpoints out that particles with small mass lead to the followingphenomena in the flotation process:

a) low particle momentum;b) heterocoagulation;(c) particle entrainment in concentrates;d) low probability of collision with a bubble; and(e) difficulty in overcoming the energy barrier between parti-

cle and particle, and particle and bubble.

While the high surface area leads to

a) a high dissolution rate in water;b) adsorption of a large quantity of chemicals;(c) rigidity of froth;d) high pulp; and

(e) slime coating.

Table 1 shows a size classification proposed by Sivamohan[18].

388 j m a t e r r e s t e c h n o l

Table 1 – Size classification according to Sivamohan [18].

Denomination Size (�m)

Super colloids <0.2Colloids <1Ultrafines <5Very fines <20Fines <100Intermediates <500Coarse >500

quality of pellets.Binders currently used by the industry act on the viscosity

In several steps of mineral processing, chemical additivesare used aiming to minimize the effects of particle size andrheology of slurry [1,5,11,19–31]. The chemical additives usedin each study, as well as their function in each case are sum-marized in Table 2.

2.3. Pelletizing of iron ore–binders

According to Moraes [7] and Lima [8], pellets are obtained byadding an appropriate amount of water to a powder; this isa fundamental factor in the formation and growth of pellets,which creates a surface tension that holds the mineral grainscohesive, thus allowing its handling.

This cohesive tension of fine particles due to water is calledneutral tension. The neutral tension, however, is not sufficientto keep cohesive grains as dense as iron minerals. Further-more, when the pellet is heated, the vaporization of wateroccurs and the pellets tend to disintegrate.

To prevent such effects, some additives, called binders, areadded to the material, aiming to:

• increase the strength of pellets before heating (greenstrength);

• prevent the collapse of pellets in the initial stages of heating,when a large volume of gas generated by water vaporizationtends to crack pellets.

Moisture and binder evenly distributed in the feeding pro-cess decisively contribute to improving the characteristics ofpellets, especially in relation to the formation of undesirableagglomerates prior to the pellet itself.

In their review paper on organic binders for iron ore pel-letizing, Eisele and Kawatra [32] use the principles suggestedby Holley [33] to categorize binders into five groups, which are:

1. Inactive film: the binder forms an adhesive layer (film) onthe particles, which promotes binding. Bonding may occurby capillary, adhesive or cohesive forces. Binding is usuallyreversible.

2. Chemical film: the binder forms a film on the surface of theparticles, which undergoes a chemical reaction and hard-ens. Binding is generally irreversible.

3. Inactive matrix: the binder forms a more or less continuousmatrix in which particles are embedded. Often, the binderis a material that needs heating to make it fluid, and thenhardens after cooling and drying. Binding may be reversible

upon heating.

. 2 0 1 3;2(4):386–391

4. Chemical matrix: the binder forms a nearly continuousmatrix, which undergoes a chemical reaction that makesit bind. Binding by this mechanism is usually irreversible.

5. Chemical reaction: the binder reacts chemically with thematerial to be agglomerated, resulting in a very strong con-nection. This occurs to some specific materials, and bindersof this type have not yet been developed for iron ore.

Bentonite is an effective, widely used binder in the iron orepelletizing process, and its low price is an important factor forits extensive use.

However, bentonite incorporates silica and alumina, whichare undesirable contaminants to pellets. Additionally, it is anatural material with variable composition depending on itsorigin. The necessary amount to obtain a suitable binder effectis very large, around 0.5% by weight, which makes handlingmore difficult and increases logistics costs.

A number of attempts to replace bentonite as a binder foriron ore agglomeration make use of organic binders. Organicbinders have advantages over bentonite, as they do not insertcontaminants, are used in smaller amounts and can be elim-inated during the pellet burning process.

The following patent documents may be cited as represen-tative of the state of the technique:

• US 3,806,414 – describes a method for using a polymericbinder on wet iron ore, either alone or with bentonite. Thebinder in this case is a graphitized copolymer of acrylicacid with a polyhydroxyl compound, cellulose derivatives,starches, sugars, etc.;

• U.S. 3,893,847 – refers to a water soluble, high molecularweight substantially linear polymer used as a binder, forexample, polyacrylamide;

• U.S. 4,767,449 – refers to a binder system composed of a poly-meric binder, for example, polyacrylamide, dispersed in anon-aqueous medium and clay, for example, bentonite;

• U.S. 4,802,914 – refers to a process comprising the applica-tion of a polymeric binder to the mineral particles in theform of a powder or non-aqueous medium in the presenceof sufficient water to bind the particles;

• U.S. 5,002,607 – refers to an agglomeration process com-prising mixing the mineral particles to particles larger than100 �m of a polymeric binder based on starch, cellulose orethylenically unsaturated monomers;

• U.S. 5,102,455 – refers to an agglomeration process com-prising the mixture of ore particles with a soluble, anionicpolymer binder, preferably a copolymer of acrylamide andsodium acrylate, also including bentonite optionally, in awet environment;

• US 5,306,327 – refers to a binder for mineral particles,based on natural starch and a soluble polymer, for exam-ple, pectin, or amide derivatives of cellulose, acrylic or vinylpolymers and others.

The state of technique is verified not to consider how toevaluate the influence of binders on the formation and on the

of the liquid phase, and the bonding of particles is obtained inany order, regardless of their size particle distribution.

j m a t e r r e s t e c h n o l . 2 0 1 3;2(4):386–391 389

Table 2 – Chemical additives used and their function in the process.

Material/process Additive Function Parameters Reference

Ironore–ultrafine/recovery

Sodiumhexametaphosphate

Dispersant

Particle size

[28]Chemical compositionStirring timeZeta potential

Coal slurry/solid–liquidseparation

Polyacrylamide FloculantFiltration

[27]Moisture content

Fe3O4 insuspension/recovery

Sodium dodecylbenzeneSurfactant used asdispersant

Zeta potential

[26]Contact anglepHAverage diameter

Hematite/grinding Acetone Grinding aidZeta potential

[25]Hardness of hematitepH

Bentonite/dispersionSodiumdodecylsulphonate

Surfactant used asdispersant

Viscosity[24]

pH

Metallicmineral/grinding

Polyacrylamide Dispersant

Size distribution

[11]ViscosityPercent solidsGrinding pathEnergy expended

Dolomite/grinding Polycarboxylic acid Grinding aid

Pulp density

[5]Pulp viscosityMulti-torqueGrinding time

Iron ore fines/selectivedispersion Sodium humate Dispersant

Percent solids

[22]Size distributionStirring timepHChemical composition

Hematite/dispersionPolyacrylic acid andpolymethacrylic acid

Suspension stability

Viscosity

[21]Zeta potentialSize distributionChemical composition

Metallicminerals/flotation

Organic polymersDepressant

Review paper [20]Dispersant flocculant

Minerals/fine grindingOrganic and inorganicdispersants

Rheology slurry control

Percent solids

[1]

Size distributionRotationViscosityTemperatureParticle shape

Iron ore pulp/dispersion Sodium hydroxide andnitric acid

Dispersion stability

Size distribution

[30]Zeta potentialPercent solidspHRotation

Quartz/flotation Polymer and surfactant Cationic interaction

Zeta potential[19]Time

RotationMinerals/grinding Organic and inorganic dispersants Grinding aid Rheological parameters [31]

Gri

SedimentationPercent solids

blewt

f

Iron ore/grinding Dispersants

Bentonite promotes the formation of ceramic bridgesetween particles, which can minimize the number of pel-

ets that collapse during firing. In case of other binders, thisffect is generally favored by their crystallization or hardening,hich creates physical connections between discrete points of

he particles.The ongoing work, conducted by this group, suggests a dif-

erent interpretation for the process.

nding aid [29]pHRheological parameters

Our proposal is to study how a colloidal wrapping filmis created from the contact of colloidal particles, particu-late material and a colloidisant agent (as compared withthe known action of a dispersant on the particles) dur-ing the formation of green pellets. This action ensures the

interstitial viscosity necessary to keep the green pellets sta-ble, as well as the stability of pellets during and afterburning.

n o l

r

390 j m a t e r r e s t e c h

Colloidisant agents, in the sense used here, are elements,mixtures or compositions with the ability to promote thedispersion of the ultrafine fraction and the formation of a col-loidal film around the iron ore particles (or agglomerates ofparticles) during pelletizing.

Such agents are polymers formed of monomers of acrylic,methacrylic, maleic, itaconic or vinylsulfonic acids, or theirderivatives (vinyl, ethoxylated, propoxylated, sulfonic or phos-phonic). Sulphide or sulfonated condensates of formaldehydewith naphthalene, phenol or melamine are also colloidisantagents. Mixtures of two or more of the monomers men-tioned above, as well as their acids and derivatives, and otherco-polymerizable monomers such as vinyl acetate are otherexamples of colloidisant agents.

Regarding the use of bentonite as a binder, or bentonitetogether with organic binders, colloidisant agents shouldprovide faster formation of green pellets, with the same orgreater strength (both green and fired pellets) and lower con-taminant content.

An advantage of this type of product is the effectivenessfor any size of pellet feed. Unlike bentonite, their efficiency isnot significantly affected by particle size distribution.

3. Conclusion

This review aimed to obtain parameters for the practical ver-ification of the action of colloidisant agents in the iron orepelletizing. Rheological parameters seem to be a good way toevaluate the effects of the application of binders which act asdispersants or vice versa.

From the studies surveyed, the rheological approach inmineral processing is strongly linked to the fine grinding step.However, the understanding of the binder action in the pel-letizing process, through rheology concepts, has not yet beenwell investigated.

Furthermore, the state of the technique does not considera way to jointly evaluate the influence of the binder on theformation mechanisms and on the quality of pellets.

In conclusion, this review allowed obtaining parameters forthe proposition of a method to evaluate the effect of bindersbefore the pelletizing step for iron ore.

Conflicts of Interest

The authors declare no conflicts of interest.

Acknowledgements

The authors thank FAPESP – Foundation for Research Supportof São Paulo and Vale for the financial support to this project,within the Call Faps Vale.

e f e r e n c e s

[1] He M, Wang Y, Forssberg E. Slurry rheology in wet ultrafinegrinding of industrial minerals: a review. Powder Technol2004;147:94–112.

. 2 0 1 3;2(4):386–391

[2] Wang YM, Forssberg E. International overview and outlookon comminution technology. FöreningenMineralteknik/Swedish Mineral Processing ResearchAssociation; 2003. p. 4–52.

[3] Possa MV. Reologia no tratamento de minérios. In: Luz AB,Sampaio JA, Franca SCA, editors. Tratamento de minérios. 4ed. Rio de Janeiro: CETEM/MCT; 2004. p. 535–70.

[4] Fuerstenau DW, Venkataraman KS, Velamakanni BV. Effect ofchemical additives on the dynamics of grinding media inwet ball mill grinding. Int J Miner Process 1985;15:251–67.

[5] Fuerstenau DW, Kapur PC, Velamakanni BV. A multi-torquemodel for the effects of dispersants and slurry viscosity onball milling. Int J Miner Process 1990;28:81–98.

[6] Pearse MJ. An overview of the use of chemical reagents inmineral processing. Miner Eng 2005;18:139–49.

[7] Moraes SL. Comparacão de desempenho de aglomeranteorgânico em relacão à bentonita na operacão de pelotizacãode concentrados de minério de ferro brasileiros de diversasprocedências. São Paulo: Escola Politécnica, UniversidadeSão Paulo; 2004. p. 80 [Tesis].

[8] Lima JRB. Estudo da carboxi-metil-celulose comoaglomerante para pelotizacão. São Paulo: Escola Politécnica,Universidade São Paulo; 1991. p. 145 [Ph.D. tesis].

[9] Cassola MS, Chaves AP. Effect of the addition of organicbinders on the behavior of iron ore pellets KONA: powderand particle. Osaka 1998;16:136–42.

[10] Ferreira EE, Brandão PRG, Klein B, Peres AEC. Reologia desuspensões minerais: uma revisão. REM: Revista Escola deMinas 2005;58(1):83–7.

[11] Kapur PC, Healy TW, Scales PJ, Boger DV, Wilson D. Role ofdispersants in kinetics and energetics of stirred ball millgrinding. Int J Miner Process 1996;47:141–52.

[12] Boger DV. Rheology and the minerals industry. Miner ProcessExtract Metallurg Rev Int J 2010;20:1–25.

[13] Shi FN, Napier-Munn TJ. Effects of slurry rheology onindustrial grinding performance. Int J Miner Process2002;65:125–40.

[14] Kawatra SK, Eisele TC. Rheological effects in grindingcircuits. Int J Miner Process 1988;22:251–9.

[15] Klimpel RR. Influence of material breakage properties andassociated slurry rheology on breakage rates in wet grindingof coal and ores in tumbling media Mills. In: Jones MJ, OblattR, editors. Reagents in the minerals industry. London:Institution of Mining and Metallurgy; 1984. p. 265–70.

[16] Mewis J, Spaull AJB. Rheology of concentrated dispersions.Adv Colloid Interface Sci 1976;6:173–200.

[17] Chaves AP. Teoria e prática do tratamento de minérios. SãoPaulo: Signus Editora; 1996.

[18] Sivamohan R. The problem of recovering very fine particlesin mineral processing – a review. Int J Miner Process1990;28:247–88.

[19] Somasundaram P, Cleverdon J. A study of polymer/surfactantinteraction at the mineral/solution interface. Colloids Surf1985;13:73–85.

[20] Bulatovic SM. Use of organic polymers in the flotation ofpolymetallic ores: a review. Miner Eng 1999;12(4):341–54.

[21] Nsib F, Ayed N, Chevalier Y. Dispersion of hematitesuspensions with sodium polymetracrylate dispersants inalkaline medium. Colloids Surf A Physicochem Eng Aspects2006;286:17–26.

[22] Mahiuddin S, Bondyopadhyay S, Baruah JN. A study on thebeneficiation of Indian iron-ore fines and slime usingchemical additives. Int J Miner Process 1989;26:285–96.

[23] Ran Q, Somasundaram P, Miao C, Liu J, Wu S, Shen J. Effect ofthe length of the side chains of comb-like copolymer

dispersants on dispersion and rheological properties ofconcentrated cement suspensions. J Colloid Interface Sci2009;336:624–33.

o l . 2

j m a t e r r e s t e c h n

[24] Permien T, Lagaly G. The rheological and colloidal propertiesof bentonite dispersions in the presence of organiccompounds v. bentonite and sodium montmorillonite andsurfactants. Clays Clay Miner 1995;43(2):229–36.

[25] Pitchai N, Nayak UB, Lahiri AK. The influence of chemicaladditives on surface charge and hardness of hematite. Int JMiner Process 1991;31:131–44.

[26] Lee H, Yamaoka S, Murayama N, Shibata J. Dispersion ofFe3O4 suspensions using sodium dodecylbenzenesulphonate as dispersant. Mater Lett 2007;61:3974–7.

[27] Moody G. The use of polyacrylamides in mineral processing.Miner Eng 1992;3(3–5):479–92.

[28] Manna M, Sasmal S, Banerjee PK, Sengupta DK. Effect of

mineral geology, mineral size and settling time onselective dispersion and separation process for recoveringiron value from iron ultra fines. Powder Technol2011;211:60–4.

0 1 3;2(4):386–391 391

[29] Vieira MG. Efeitos da reologia de polpa e avaliacão dedispersantes como aditivos de moagem na remoagem de umconcentrado de minério de ferro. Minas Gerais: UFMG; 2011.p. 151 [Ph.D. Tesis] Curso de Pós-Graduacão em EngenhariaMetalúrgica e de Minas, Universidade Federal de MinasGerais, Minas Gerais.

[30] Abro MI, Pathan AG, Andreas B, Mallah AH. Effects of variousparameters on the dispersion of ultra fine iron ore slurry.Part-1. Pak J Anal Environ Chem 2009;10(1–2):34–8.

[31] Klimpel RR. The selection of wet grinding chemical additivesbased on slurry rheology control. Powder Technol1999;105:430–5.

[32] Eisele TC, Kawatra SK. A review of binders in iron ore

pelletization. Miner Process Extr Metall Rev 2003;24:1–90.

[33] Holley CA. Agglomeration: the state of the art. In:Symposium on fine powder processing 109th AIME annualmeeting. 1980.