Factors Affecting Selective Flocculation

Embed Size (px)

Citation preview

  • 8/3/2019 Factors Affecting Selective Flocculation

    1/4

    Vol. 10 Special ssue Trans. NonferrousMet. Soc. China Jun. 2000Article ffi: lOO3-6326(2000)Sl-OOO8-04

    Selective flocculation of finesP. Somasundaran,V. RunkanaLangmuir Center or Colloids and Interfaces,ColumbiaUniwrsity, New York. NYlOO27. USA

    Abstract: A numberof factors hat affect selective kJCcu1ationf fineshavebeen dentifiedand the effect of ~e of pa-rameters n p~ behaviorhasbeenexplainedwith the help of a few case tudies. Experimentswith francolite-montmo-rillonite and rancolite-palygorskite ixtures ndicated hat francolite ecoverydepends n pH and the type of dispersant.The resultsshowed hat removalof multivalent ionic species n clay mineral surfaces eems o enhance locculation. andseparation fficiency ncreases s Caz+ ons are removed rom the surface. When chalcopyriteand quartz are present o-gether. it is howevernecessaryo clean he fIocsobtained o removeentrapped uartz.Key 'WordS: flocculation; clay minerals; p~phate minerals; dispersant Document code: A

    INTRODUCTION 1.1 Basic principles and mechanisms of separationThe basic principles of selective flocculation are

    actually similar to those of the conventional selectiveflotation process. First, the individual mineral parti-cles need to be liberated and dispersed, then a reagentis required that can selectively adsorb onto one or sev-eral of the mineral surfaces. So, the two main criteria(or selective flocculation are dispersion of the fines andselective polymer/reagent adsorption.

    When particles are in suspension, they collide ei-ther due to the Brownian motion or due to externalforces induced by agitation, magnetic field etc. Theprobability of aggregation during such collisions is de-termined by the nature of interactions between theparticles. Attractive interactions arise due to London-Vander Waals type of forces whereas repulsive inter-actions arise generally due to the electrostatic forces.There is always a net total energy of interaction and itis necessary that the following conditions are met sothat selective aggregation can take place:

    1) All types of particles should carry the samecharge so that there will be no heterocoagulation be-tween them. Repulsive energy should be larger thanthe energy of attraction between different particles.

    2) Charge on the particles to be aggregatedshould be such that repulsive energy between themshould be less than that of the attractive energy.

    Floc formation takes place by either of the fol-lowing three mechanisms:1) Reduction of electrostatic repulsion betweenparticles;2) Formation of polymer bridges between parti-

    As the earth's resources are being consumedsteadily and as demand for mineral products continuesto increase, the need to utilize -low-grade ores moreefficiently has become more severe. This indeed re-sults in the generation of a large quantity of fine parti-cles during mining and processing of such ores. Largeamounts of fine mineral values are discarded currentlyas techniques to recover ultra-fine minerals are not ad-equate. Conventional processing techniques such asflotation are not efficient to process fines due to theintimate mixing of values and gangue materials. Thisis especially true in the case of fines in the sub-micronrange. Also, these fines cannot be disposed off easilyas it results in environmental hazards and in some cas-es, waste of useful land. -Thus, there is a great need to develop processesto utilize the fine ores and selective flocculation hasshown some promise and potential in this regard. Theobjective of the present communication is to elucidatethe basic principles and mechanisms of selective floc-culation, identify the critical process parameters, pre-sent results obtained for materials of industrial impor-tance and suggest directions for future research.2 SELECflVE FLOCCULATION

    This process nvolves selectively flocculating thedesirable!undesirablemineral from a mixture of min-erals. Though this appearssimple, it is quite complexin nature. Selective flocculation has to always dealwith a binary or even a multicomponent mixture ofmaterials. Also, long chain polymers are normallyused as flocculants. So, it is necessary o understandnot only the interactions between particles, polymersand the solvent but also the interactions between dif-ferent particles in the pulp. The surface chemistry ofdifferent materials in the presenceof each other be-comesan important factor in this situation.

    cles; \3) ion-exchange eactions between polymers andparticle surfaces. .Selectivity in flocculation can be achievedby anyone or combination of the following treatments:1) Altering the surface potential of differentminerals;2) Incorporating specific functional groups into

  • 8/3/2019 Factors Affecting Selective Flocculation

    2/4

    Vol. 10 Special ssue Selective flocculation of fines 9with the dissolvedspecies.Experimental results obtained from the studieson selective locculation of synthetic mixtures of chal-copyrite and pentlandite using PEa and DPG are pre-sented n Fig.!. It can be observed rom Fig.! thatflocculation of the mixture is enhanced by the pres-ence of DPG and PEa together. particularly underacidic pH conditions.

    the polymer chain;3) Controlling flocculation! polymer-particle in-teraction time;4) Selectivepolymer adsorption;5) Coating impurities or specific ions on solidsurfaces.2.2 Important processparametersSome of the important factors that affect selec-tive flocculation are1) size distribution of colloidal particles;2) shape and surface heterogeneity of particles;3) concentration, molecular weight distributionand charge density of the polymer;4) functional groups present on the polymerchain;5) suspension H,' temperature, density ~d vis.cosity;6) dissolved onic species n solution;7) speedof agi ation and type of impeller.3 CASE STUDIFS

    We now present a few examplesof selective loc-culation in which the effect of some of the factorsmentioned above s discussed n detail.3.1 Selective locculation or sulfidesAs indicated earlier, selective flocculation of amixture of minerals is significantly different from thatof a single mineral. One of the reasons or this is thedissolution, precipitation! readsorption of mineralspecies rom one mineral onto another. When mineralfines aresuspendedn solution, species uchasCu2 ,N? +, Fe2 and 52- present n the surface f a min-eral can dissolveand then precipitate! readsorbon an-other mineral surface. I t has been reported thatspecifically adsorbing ons can change and/or even re-verse the zeta potential of a mineral and thereby causeheterocoaguiation s well as enhancedpolymer floccu-lation[1-3]. Acar and SomasundaraJ4] studied sepa-ration of chalcopyrite and pentlandite from a mixtureof these minerals using polymers such as polyacry-lamide (PAM) and polyethylene oxide (PEO). Floc-culation was observed both in the presenceand ab-senceof the polymer. Though the presenceof PEaenhanced the efficiency of overall flocculation, theseparationefficiency was not significant.Electron spectroscopy or chemical analysis (ES-CA ) and electrophoretic mobility measurementsshowed that nickel ions present on the chalcopyritesurface and copper ions present on the pentlanditesurface causenonselectiveadsorption of polymer. Theamount of dissolved specieswas significant especiallyunder the acidic pH conditions. In order to overcomethis problem diphenylguanidine (DPG) was intro-duced nto the system so that it can form complexes

    3. 2 Separation of phosphate minerals from claymineralsFlorida phosphatic clay is disposed off in enor-mous quantities though it is rich in phosphateminer-als and causes serious environmental hazards. Thewaste is in the form of fine slimes and because f thisthe clay minerals such as montmorillonite and paly-gorskite are intricately mixed with the ph~phatemineral, francolite. Traditional separation processessuch as flotation are extremely inefficient because heslimes are in the micron and sub-micron size range.Another problem is the presenceof organic matter andcementing materials such as oxides and hydroxides ofAI and Fe. Andersen and Somasundaran[S] tudiedthis separation problem by preparing synthetic mix-tures of clay and ph~phate minerals and using thenaturally occurring slimes. After initial screening ex-periments. polyacrylic acid with an averagemolecularmassof 3 million was found to be best suited for poly-mer flocculation. Sodium silicate and sodiumtripolyph~phate were used as dispersants. Electroki-netic measurements ndicated that all the pure miner-als as well as the natural slimes are negatively chargedin the entire pH range (4-11) studied.Experiments with francolite-montmorillonite(Fig.2) and francolite-palygorskite (Fig.3) mixturesindicated that francolite recovery dependson pH andthe type of dispersant. For both the mixtures, recov-ery decreased t high pH, where it is likely that elec-trostatic repulsion is inhibiting polymer adsorptionand flocculation. The behavior of the two systemswith respect o francolite recovery at lower pH values

  • 8/3/2019 Factors Affecting Selective Flocculation

    3/4

    Jun. 20000was quite different. In the cue of montmorillonite,gel formation was observed after polymer addition,resulting in a decreasen the recovery. In the caseofpalygorskite, recovery ncreasedwith decreasingpH.The floc grade in both systems s hardly affected bypH. The difference between the two systems can beexplained by examining the differences in .the crystalstructure of the two clay minerals. Due to the edgeface charge characteristics, the plate-shapedmontmo-rillonite particles can form a card-house structure insuspension.This prevents the clay particles as well asthe francolite particles from settling which results inlow recovery of francolite. The needle-shapedpaly-gorskite crystals cannot form such an expandedstruc-ture in the suspension. So these particles flocculateand setde together with francolite. For both the sys-tems, maximum separation occurred in the pH rangeof 9-9.5 with S(x{iumsilicate as the dispersant.

    Trans. Nonferrous Met. Soc. China. - -~

    080 8#

    1 I .~'.4 J60.0

    II~ 40~c.. 20

    1".8#

    1

    1:llobJt.:F~4 x 10-' PM2~ 8OIids 0 66 0 I a6 a 0 A .. IC6 1t--

    I)

    55 .20

    ~145II. I

    " J-- ~. i Lo_~! """""'~"'I"""'b1:1~.:F~. ~r~s9.3-9.S :I ~~ 10~.'.PAA :I 2~ IOIi.dI . ..0 20 40 ~ .. 100'

    Ca bdiII8 011 a,./ ~IO.~---~ -~,~-

    -c 111.2Fig.4 Effect of the presenceof exchangeablecalcium and sodium on separationof francolite-montmorillonite mixtures by poiyacrylic acid

    3:

    Fig.1 Separation esults for francolite-montmorillonite mixtures using polyacrylic acid9Or1:1 Palll~:F~~.A -- ",.04 x 10-iPAA 0-101lll~..2-. dids oA-O. 5 1li STPP/..# 80

    'toI 70"1 60I&. so

    0.8M

    6 1t--04 OJ~

    --:l '0A_~

    . ~~40' _~~_~-A. '-!8.S 9.0 9.S 10.0 10.S

    pH

    0.2

    Fig.! Effect of the presenceof exchangeablecalcium and sodium on separationof francolite-palygorskite mixtures by polyacrylic acidFig.) Separation esults for francolite-palygorskite mixtures using polyacrylic acid Selective locculation of natural phosphaticslimeswas attempted next. Sedimentation experiments ndi-cated that the as-received natural slimes flocculatedand settled at pH 7 within one-half hour. But. selec-tive flocculation experiments. under the optimumconditions obtained for synthetic mixtures. did notresult in any specific separation of francolite. The

    The separation achieved in the above experi-ments was far from the maximum that can beachieved heoretically. Pure francolite contains 32 %P2o,. The feed had 16% P2o, while the best thatcould be achieved was 28% in the presenceof mont-

    0.6to

    -J A" o-101D1 $/1I . A'~ .A-o..51D1s:w/IIO'.0 7.5 8.0 8.5 9.0 9.5 10.0

    pH

  • 8/3/2019 Factors Affecting Selective Flocculation

    4/4

    Vol. 10 Special ssue Selective flocculation of fines 'n-main reason or this appeared o be the presence f ce-menting materialssuch as organic matter and inorgan-ic (mainly AI and Fe) hydroxide precipitates whichinhibit dispersionof particles. AI and Fe were foundto be 10 times more abundant in natural slimes thanin pureclayminerals.The organicand norganicQ18t-ter causesaggregation of particles. In addition, theAI and Fe compoundsmight be causing non-selectivepolymer adsorption also.3. Beneficiation of mineral slimesThe type of functional group on a polymer chaincan also modify the behavior of flocculation. Srestyand Somasundaran[6] tudied selective flocculation ofchalcopyrite-quartzsystem n the presenceof hydrox-ypropylcellulose xanthate. Hydroxypropylcellulosewas modified to incorporate the xanthate functionalgroup, as t is known that xanthates adsorb selective-ly on heavy minerals such as chalcopyrite, galena andsphalerite. Fig. 6 shows the variation of percent solidssettled as a function of xanthate concentration. It canbe observed that chalcopyrite flocculates and settlesvery well whereasquartz stays suspended n solution.When the two minerals were present together, it washowever n~ to clean the flocs obtained to re-move entrapped quartz.

    can be overcome by introducing a complexing agentinto the system.A number of ,fa~tors hat affect selective loccula-tion of fines have been identified and the effect ofsomeof theSeparameterson processbehavior has beenexplained in this paper with the help of a few casestudies.It has been found[7] that polymer: conformationat the solid-liquid interface has a significant impact onflocculation. It is necessary o understand how theconformation changes depending on the mineral sur-face heterogeneity and take advantageof that for se-lective adsorption. Also, there is a need to developnew processing echniques o remove undesirable ons(for example, Ca2+ ons in the caseof Florida phos-phatic slimes) present on the mineral surfaces n mul-ticomponent mixtures of minerals.Hydrodynamics will playa major role in the caseof large scale ndustrial separationoperations. Howev-er, very little research has been done in this direc-tion. It needs to be seen as to how existing knowl-edgeon mixing of fluids can be extended for particu-late suspensions.Even though a number of mathematical modelshave been proposed n the literature, very few takepolymer adsorption into acCbunt. It is important toincorporate knowledge on all aspectsof flocculation inorder to arrive at meaningful conclusions.ACKNOWLEDGMENTSThe authors acknowledge the National ScienceFoundation for its financial support (NSF/EEC-9804618).~~~~~

    REFERENCES

    300100 200Xanthate COlIC. Dry dids basis)/10-'

    Fig. 6 Percentageof chalcopyri e fines and quartzfines settled as a function of concentration ofhydroxypropylcellulose xanthate

    (Reagentizing time, 30 s; settling time. 45 s)4 CONCLUSIONS

    Selective locculation has been found to be a fea-sible technique for the separation of sulfides, ox.idesand phosphatic minerals. However, dissolution ofmineral specieshas a significant impact on the selec-tivity of polymer adsorption. This problem however,

    [1] Critchley K and ewitS R. The effect f Cu' + ionsonzeta potential of quart~ (J]. Trans Inst Min Metall (SectC: Miner Process Extr Metall) , 1979,88: C57-59.

    [2] DrzYmala J and Fuerstenau D W. Selective flocculation ofhematite in the hematite-quartz-ferric ion-poiyacrylic acidsystem. 1. Activation and deactivation of quartz (J]. IntJ Miner Process, 1981, 8: 265-277.

    [3] Critchley J K and Straker P. Flotation of nickel sulphideand zeta potentials in a nickel (2)-xanthate system [J].Trans Inst Min Metall (Sect C: Miner P~ Extr Met-all), 1981, 80: 44-45.

    [ 4 ] Acar S and Somasundaran P. Flocculation of sulfides andthe role of a OOffiplexingagent in it (J]. Int J Miner Pro-cess, 1989.. 27: 111-123.

    [5] Andersen B and Somasundaran P. Mechanisms determin-ing separation of phO5phatic clay waste by selective floccu-lation (J]. Miner and Metall Process. 1993: 200-205.

    [ 6 ] Sresty G C and Somasundaran P. Selective flocculation ofsynthetic mineral mixtures using modified polymers [J].Int J Miner Process, 1980, 6: 303-320.

    [7] Yu X and Somasundaran P. Role of 'polymer conforma-tion in int~rticle-bridging dominated flocculation [J].J Colloid Interface Sci, 1996, 177: 283-287.