Bridging Theory And Practice In Iron Ore Flotationyunus.hacettepe.edu.tr/~obicak/MAD454/Iron Ore_Iwasaki.pdf · 2017-03-06 · BRIDGING THEORY AND PRACTICE IN IRON ORE FLOTATION 179

Chapter 20

BRIDGING THEORY A N D PRACTICE I N IRON O R E FLOTATION

I. Iwasaki

Mineral Resources Research Center University of Minnesota Minneapolis. Minnesota

Abstract. Iron ores may be upgraded by flotation of iron oxides or siliceous gangue using either cationic or anionic collectors. The electrostatic mechanism of adsorption can account fo r the complex chemistry involved. In industrial practice, only cationic silica flotation is used a t present in North America. Some of the seemingly paradoxical phenomena of cationic silica flotation behaviors in meeting the demand for low silica concentrates are reviewed.

Low-grade iron ores that occur in vast quantit ies in North America are, in general, intimately associated with silicious gangue and require f ine grinding for liberation. Earlier, prompted by the dwindling reserves of direct shipping and easily concentratable wash ores in the Lake Superior and Birmingham, Alabama districts, the research effor ts were directed towards the development of flotation technology on oxidized iron ores as a primary means of concentration. The principles and industrial practice of iron ore flotation have been reviewed extensively (Houot, 1983; Iwasaki, 1983; Nummela and Iwasaki, 1986).

Due to a n excess capacity in the world a t present, the flotation processing of low-grade hematit ic iron ores is no longer an attractive target. A major interest has recently been directed towards upgrading magnetic taconite concentrates which, when ground to 80-90% -44 pm (-325 mesh) or even f iner , still contain locked siliceous middlings to the extent of 5-9% SiOz that cannot be removed by repeated magnetic or hydraulic separation. The development of low silica concentrates by cationic silica flotation, therefore, is receiving increasing attention by the iron ore industry, particularly for the production of fluxed pellets, to improve the productivity and energy efficiency of blast furnaces. The present practice demands a uniform silica product typically in the range of 4 to 5% for providing sufficient slag volume for sul fur control. Direct reduction- electric furnace smelting may be economically competitive with the conventional blast furnace - BOF steelmaking under certain specific circumstances. Feed materials for direct reduction should preferably contain less than 2% SiO,. Future technologies of direct iron- and steelmaking may impose yet lower gangue requirements, either in concentrates or in pellets.

In this review, an emphasis was placed on some of the parameters that influence the flotation removal of

siliceous gangue f rom hematite and magnetite concentrates in order to meet the increasingly critical demands of the steel industry.

CHOICE O F FLOTATION METHODS

Iron ores may be upgraded by flotation of either iron oxides o r siliceous gangue using either a cationic or anionic collector as illustrated in Table 1. Each

Table 1. Classification of Iron Ore Flotation Methods

R Paraffin chain containing 12 - 18 carbon.

method has its own optimum operating conditions, and well illustrates the complexity of the chemistry involved. The theoretical foundation leading to the selectivity of adsorption of flotation reagents on oxide and silicate minerals has been a subject of intensive research over the years and a number of comprehensive reviews are available (Aplan a n d Fuerstenau, 1962; Fuerstenau, 1970; Fuerstenau a n d Fuerstenau, 1982; Fuerstenau and Palmer, 1976; Smith and Akhtar, 1976). I t is now clearly established that for oxide minerals, hydrogen and hydroxyl ions are the potential determining ions and control the surface charge, which, in turn, governs the selectivity of collector adsorption. The selectivity of flotation separation may then be related to the d i f ference in the points-of-zero-charge (PZC) of the constituent minerals in ores. An extensive list of the PZC's of solid oxides and hydroxides is available in which the effects of cationic charge, hydration, impurities and nonstoichiometry on PZC's are also discussed (Parks, 1965). Lists of the PZC's of silicate minerals and their relations to flotation are also available (Fuerstenau and Palmer, 1976; Fuerstenau and Raghavan, 1978).

Siliceous Gangue Flotation

RCOO ' + Ca++ (pH 11-12)

Lab pH-6 RNH; (plant pH 611 )

Anionic Collector

CaWonic Collector

The iron ore flotation methods interpreted in this manner may be theoretically simple, yet they are

Iron Oxide Flotation

1. RCOO ' (pH 6-91 2. RSO,',RSO,' (pH 2-4) 3. RS0,- + RCOO (pH- 5.5)

RNH,+ + F- (pH 24)

178 ADVANCES IN COAL AND MINERAL PROCESSING USING FLOTATION

exceedingly complex systems in practice, particularly when ever tightening specifications for the concentrate grades are to be met. The selectivity of flotation separation is influenced by a complex combination of mineralogical, chemical, physical and engineering parameters. The choice of a flotation method for a given iron ore, therefore, involves a careful assessment of these parameters.

Of foremost importance is the mineralogical characteristics of the ore. Hard martite ores consisting of approximately equal amounts of hematite and siliceous gangue may be upgraded by floating either iron oxides or silica. For specularite ores, the anionic flotation of the hematite would be preferable for taking advantage of its readily floatable surface propcrty (Johnson and Bjorne, 1964). A hydrophobic coating on iron oxides, however, causes operational problems in pelletizing. Moisturc accumulates in the drying stage and also lowers the decrepitation temperature, both leading to difficult ies in the induration furnace. For magnetic taconite concentrates, there is a small amount of locked siliceous gangue and, therefore, the cationic silica flotation would be the method of choice (DeVaney, 1943; Scott. Richardson and Arbiter, 1945). Limonitic iron ores with porous and easily abrading surfaces are extremely di f f icul t to process. In fact , a flotation scheme optimized on a given oxidized ore sample may prove to be totally ineffective to other ores or cven to a sample secured only a short distance away within the same mine (Beebe, 1965).

Those processes that require highly acidic pulps have serious equipment corrosion problems and also the plant ef f luent becomes of environmental concern. A positive control of concentrate grades is a n important parameter in meeting the product specifications for iron ores. Cationic silica flotation is amenable to automatic control in conjunction with neutron activation analysis for silica, but it may be di f f icul t with anionic silica flotation. Furthermore, the highly alkaline pH of the anionic silica flotation may be economically unattractive.

Two commercial plants processing oxidized iron ores in North America have used cationic silica flotation which illustrates the parameters examined in the choice of this flotation process. At the Sept-Iles plant, Quebec, now closed, cationic silica flotation was favored because desliming was not necessary and the method was most adaptable to the Knob Lake ore. The three anionic flotation processes were considered but rejected for the following reasons: petroleum sulfonate flotation of iron oxides because of sensitivity to slimes; fa t ty acid flotation of iron oxides because of reagent costs, difficult ies in f ro th handling and decrepitation of the green balls in the pelletizing process; and calcium activated anionic silica flotation because of reagent costs (Bunge, et al., 1977).

In the development of the Tilden, Michigan, flowsheet, both calcium activated anionic silica flotation and cationic silica flotation were evaluated. Cationic silica flotation was chosen because of the reagent cost for maintaining the pulp pH near 10.5 with amine as against 11.5 in the anionic silica flotation. Another factor in favor of the amine was the ability to exercise grade control via the amount of cationic collector added (Nummela and Iwasaki, 1986).

SELECTIVITY O F FLOTATION SEPARATION

Figure 1 shows the electrophoretic mobilities and the flotation characteristics with dodecylammonium chloride and sodium dodecylsulfate on quartz, magnetite and cummingtonite [(Fe,Mg),Si80,,(OH),1, which are the three major constituent minerals in magnetic taconite of the East Mesabi Range, Minnesota. It is apparent in the figure that the floatabilities of the three minerals using these collectors correlate well with their surface electrical conditions and suggest that the selectivities of flotation separation may be related to the difference in the PZC's of the constituent minerals. The intermediate behavior of cummingtonite indicates the d i f f icul ty of achieving high selectivity of separation by the cationic silica flotation either by insufficient removal of siliceous gangue or by excessive loss of the iron units.

F IGURE 1. a ) Electrophoretic Mobility of Magnetite (M), Cummingtonite (C), and Quar tz (Q) as a Function of pH. b) Flotation of Magnetite, Cummingtonite, and Quartz with lo-' M Dodecylammonium Chloride. c) Flotation of Magnetite, Cummingtonite, and Quartz with lo-' M Sodium Dodecylsulfate (Iwasaki, Cooke and Choi, 1961).

The three iron oxides, hematite, goethite and magnetite, have their PZC's within a narrow pH range of 6.5 to 6.7, a n d hence their flotation characteristics of well-sized and hard varieties are indeed qui te similar both with the cationic and anionic collectors (Iwasaki, Cooke and Kim, 1962). The influence of pH on the selectivity of flotation separation of iron ores may be demonstrated on an artif icial mixture of well-sized goethite and quartz with the cationic and anionic colle.ctors in terms of selective indices (Gaudin, l957), as shown in Figure 2. With dodecylammonium chloride.

BRIDGING THEORY AND PRACTICE IN IRON ORE FLOTATION 179

Sodium dodecyl sulate 0 Dodecylammonium chloride

F IGURE 2. Selectivity Indices as a Function of pH on a Mixture of Quartz and Goethite, with M of Sodium Dodecy l Sulfate and of Dodecylammonium Chloride (Iwasaki, Cooke and Colombo, 1960).

the maximum selectivity of separation was obtained in the range of pH 6 to 7, whereas with sodium dodecylsulfate, it was approximately a t pH 2 in good agreement with the flotation characteristics of the single mineral systems presented in Figure 1.

T h e flotation separation can then be achieved in the pH range bracketed by the PZC's of the two minerals, where they a re oppositely charged. However, this pH range also corresponds to the condition where heterocoagulation of the two minerals can occur and the flotation separation becomes more di f f icul t as the particle size becomes finer. Usui and Takeda (1983, 1983a) made a detailed study of the flotation behaviors of hematite and quar tz in three size fractions (149/210, 53/44 and -10 pm) both individually and as equal weight mixtures. Figure 3 shows the flotation results as a funct ion of the collector concentration in acid pulps. When the two minerals were floated individually, all three size fractions of quar tz particles responded to flotation similarly while very little hematite floated as expected. When the equal weight mixtures of the respective size fractions of quar tz and hematite were floated, an excellent separation was obtained on the coarsest mixture and a reasonably good separation on the intermediate size mixture near M, but the -10 pm mixture showed very little selectivity indicating that heterocoagulation of hematite and quar tz overshadowed the di f ference in the adsorption behaviors of the cationic collector.

In a moderately alkaline pulp of near pH 10, both hematite and quar tz a re negatively charged and adsorb cationic collectors. Since the amines a re not particularly selective for the flotation separation of siliceous gangue a t this pH, a suitable hematite depressant is needed for

(A) Single

Flotation

8

Concentration of DAA, M

FIGURE 3. Effect of Dodecylammonium Acetate Concentration on the Flotation Recoveries of Quartz (open symbols) and Hematite (filled symbols) Showing the Effect of Particle Sizes. (A) Single Mineral Flotation, (B) Flotation f rom Equal Weight Mixtures of Respective Sizes ( a f t e r Usui and Takeda, 1983 and 1983a).

satisfactory separation. Starches and their derivatives are commonly used for this purpose. Nonionic polyacrylamide also appears to ac t as a depressant for hematite (Somasundaran and Lee, 1982). In fact , the earthy hematite ores a t the Sept-Iles plant were processed a t a pH of 10 to 10.5 using a diamine as a collector and a wheat dextrine as a depressant fo r iron oxides (Bunge e t al., 1977). The pulp pH of the cationic silica flotation at the Tilden concentrator is also reported to be in the range of 10 to 11 because of the caustic soda carried over f rom the selective desliming step (Villar and Dawe, 1975). Apparently, a combined use of a moderately alkaline pH and a starch depressant is preferable to the simple control of pulp pH between the PZC's of hematite and quar tz in order to circumvent heterocoagulation.

Coburn (1985) reported, however, that, in the cationic flotation of finely ground oxidized iron ores, f ree iron oxide particles were recovered in the f ro th due to co-flocculation (or heterocoagulation) of iron oxide to quartz-covered bubbles even at a pH near 10. Small additions of decyl alcohol, dodecylsulfatc or EDTA with dodecylsulfate was noted to increase double layer repulsion between the particles thereby diminishing the co-flocculation.


E F F E C T O F W A T E R Q U A L I T Y

T h e impor tance of water qua l i ty has long been recognized in t h e desl iming and f lo ta t ion of i ron ores. In f resh water supply, Na+, C a f z , a n d ~ g + ' a r e most commonly f o u n d , A I + ~ , ~ e + ' a n d ~ e + ~ may occur in ac id ic water , whereas an ions inc lude C03=, SiO,' a n d sometimes SO,' when su l f ides a n d gypsum a r e present. Oxidized i ron ores ground i n dist i l led water typical ly release t o 10'' M of the a lka l ine e a r t h ions a t near neut ra l pH. These ions cause not only t h e ac t iva t ion of siliceous gangue in the f a t t y ac id f lo ta t ion of i ron oxides, b u t also t h e depression of s i l iceous gangue in t h e ca t ion ic silica f lo ta t ion , thereby contamina t ing t h e iron concentrates in both methods. For this reason, thc select ivi ty o f t h c a n i o n i c f lo ta t ion of i ron oxides is known t o improve by using c i ther softened watcr o r ion cxchange rcsins in pulps.

Ca t ion ic si l ica f lo ta t ion is more to le ran t o f water qua l i ty t h a n the f a t t y ac id f lo ta t ion of i ron oxides, bu t a n ex tens ive recycle of watcr in f lo ta t ion p lan ts m a y lead t o a b u i l d u p of soluble salts, f lotat ion reagents, p roducts of vcgetat ion dccomposit ion, a n d cvcn micro- a n d macrobiological specics in tai l ing ponds. Developmental work in t h e labora tory a n d pi lot plant is o f t e n car r ied o u t wi th dist i l led or t a p water total ly unre la ted in q u a l i t y t o the eventua l commercial opera t ion . A f i n a l eva lua t ion w i t h plant water . therefore. is essential in the deve lopment of a f lo ta t ion f lowsheet .

F r o m adsorp t ion mcasurernents, s t reaming potential measurements a n d f loccula t ion tests, i t h a s been postulated t h a t a t near ncut ra l pH, ca lc ium ion is adsorbed a s h y d r a t c d ~ a + ' , whereas a t pH 1 1 a s a t iydroxy complex C ~ O H + (Iwasaki , et . al., 1980). By cor re la t ing wi th the f lo ta t ion bchaviors, t h e f lo ta t ion of Ca-ac t iva ted q u a r t z wi th f a t t y ac id collectors was postulated to involve C a O H f a s t h e e f fec t ive f o r m f o r ac t iva t ion (Clark a n d Cooke, 1968). T h e h y d r o x y complex C a O H + appears to play a role that d i f f e r s f r o m hydra ted C a + 2 in t h e ca t ion ic si l ica f lo ta t ion a s well, a s shown in F i g u r e 4. A t neut ra l pH, the recovcry was unaf fec ted by increased Ca+' a n d Mg+' concentrat ions.

Electrolyte Concentration, M

F I G U R E 4. F lo ta t ion of Q u a r t z a s Funct ions of Calc ium a n d Magnesium Chlor idc Concent ra t ions a t pH 7 a n d 10.5 (Iwasaki , Smith, L ipp , a n d Sato, 1980).

Ear l ie r , O n o d a a n d Fuers tenau (1964) reported t h a t in t h e f lo ta t ion of q u a r t z w i t h dodecylammonium acetate, inorganic ions depress f lo ta t ion of q u a r t z th rough ionic compet i t ion a t low collector concentrat ions, b u t they have l i t t l e e f f e c t a t high collector concent ra t ions where t h e collector is s trongly adsorbed through a hydrocarbon c h a i n in te rac t ion . At p H 10.5, however, the f lo ta t ion recovery decreased g r a d u a l l y wi th increas ing concent ra t ion of ca lc ium ion imply ing t h a t s trongly adsorbed C a O H + m a y have inf luenced the adsorp t ion of the collector.

T h e adverse e f f e c t o f CaOH+ can be remedied by the use of complexing agents, a s shown in F i g u r e 5a. I n t h e presence of M CaCI,, t h e f lo ta t ion recovery was lowered to a b o u t 80%. A n a d d i t i o n of a small a m o u n t of s o d i u m t r i p o l y p h o s p h a t e ( S T P P ) o r s o d i u m hexametaphosphate (SHMP) f o r m e d precipi tates, which resulted in a m a r k e d d r o p in recovery t o near zero. T h i s phenomenon is a t t r i b u t a b l e to t h e heterocoagulat ion of q u a r t z with t h e precipi tatcs, result ing in sl imc coating. At h igher levcls of the i r addi t ion , near ly f u l l recovcries were ob ta ined . In t h e casc of STPP, f u l l recovery was ob ta ined a t the stoichiometric a m o u n t suggest ing t h a t near ly a l l ca lc ium a n d tr ipolyphosphate h a d becn removed f r o m the q u a r t z sur face . SHMP.

Percent of Stoichiometric Amount

F I G U R E 5. Flotat ion Rccovery of Q u a r t z a t M Dodecylammonium Aceta te a s Funct ions of Stoichiometric Amounts of Complexing Agents (Heerema, Lipp, a n d Iwasaki , 1983; Iwasaki , Smith, Lipp, a n d Sato, 1980).

however, d i d no t reach fu l l recovery un t i l 125% of the s to ich iomet r ic amount . T h e f lo ta t ion recovery curve f o r E D T A shows v i r tua l ly f u l l recovery a t a s low a s 25% of t h c s to ich iomet r ic amount . Apparen t ly , this complexing agent not only removed the ca lc ium ion f r o m q u a r t z sur face , b u t also was a t t rac ted t o the ca lc ium ion o n t h e q u a r t z s u r f a c e a n d a i d e d t h e f lo ta t ion of q u a r t z wi th t h e ca t ion ic collector.

These observa t ions suggest t h a t t h e type o f complexing agents a n d t h e manner in which these

BRIDGING THEORY AND PRACTICE IN IRON ORE FLOTATION

reagents may be used should be carefully assessed. Perhaps chemical precipitation, say with soda ash, followed by high solid scrubbing or ultrasonic treatment, as will be discussed under selective desliming, may alleviate the adverse effects of calcium ion without introducing environmentally sensitive chemicals to the pulp.

Magnesium ion is also found in iron ore pulps though not in as high a concentration as calcium ion, and it appears to a f f ec t the flotation behavior of quar tz differently. With magnesium ion, the flotation recovery dropped abruptly to nil above M, as shown in Figure 4. In the region where the flotation recovery was nil, white precipitates were observed. Such a flotation behavior was attributed to a slime coating caused by the heterocoagulation of magnesium hydroxide precipitates on the quar tz surfaces.

When the same complexing agents, EDTA, STPP and SHMP, were added in excess of stoichiometric amounts, all the complexing agents restored the flotation of quartz particles, as seen in Figure 5b. Apparently, the magnesium hydroxide precipitates were dissolved and the quar tz surfaces were made accessible to the collector ion. Sodium silicate, however, is ineffective in restoring the floatabili ty when the magnesium hydroxide coating is present.

EFFECT O F PULP TEMPERATURE

The effect of temperature on the fa t ty acid flotation of oxidized iron ores was investigated by Cooke et al. (1960). The selectivity indices of the flotation separation on a deslimed ore increased linearly with the pulp temperature, and even a n undeslimed sample was shown to produce reasonably good separation a t 70°C. At room temperature, the undeslimed ore was totally unresponsive to collection of the iron oxide minerals with fa t ty acids, irrespective of the addition levels of the collector. The increase in selectivity index, when iron ores are floated a t higher temperatures, may be interpreted on the basis of the floatabilities of hematite and of calcium-activated quar tz which are, respectively, positively and negatively temperature-dependent a t pH 6. The temperature-dependence may then be due either to diminished calcium adsorption, or to a lower calcium-soap solubility.

The hot flotation, that was in practice a t the Republic Mine in Michigan, involved regrinding of the specularite concentrate floated with a fa t ty acid collector followed by re-conditioning near boiling point wi th steam injection. T h e re-conditioning a t a n elevated temperature rejected the silicates in the subsequent flotation.

The information on the effect of temperature on the cationic silica flotation is sparse due presumably to the relatively minor ef fect this parameter has to the selectivity of separation. Hruz (1954) investigated the ef fect of temperature on the flotation of an oxidized iron ore using primary amine collectors with ten to eighteen carbons. All the collectors showed improved effectiveness with a n increase in temperature and the amounts of the collectors required in each case to float 90% of the quar tz a t 51°C were reduced by half . In the hot flotation of a deslimed oxidized iron ore, the concentrate grade was improved, but a t the expense of the iron recovery, implying that the effect of pulp

temperature on the selectivity of separation was minor.

In a pilot plant study on the cationic silica flotation of a deslimed oxidized iron ore a t the MRRC in the mid 19603, the selectivity of separation was noted to suffer progressively as the pulp temperature fell f rom fall to winter and virtually no separation was obtained when the temperature reached near 5°C. Such a flotation behavior was thought to be related to the solubility of the collector a t lower temperatures. Heating of the process water to room temperature by steam injection restored the effectiveness of the flotation separation.

Conventional Desliming

In the flotation of oxidized iron ores, desliming is essential to successful separation, irrespective of whether the iron oxide or siliceous gangue minerals are floated, although anionic silica flotation appears to be less sensitive to the presence of slimes. The detrimental effect of slimes is twofold. The presence of slimes leads to a high reagent consumption because of high specific surface. When clays are present and when cationic collectors are used, abstraction of the collector by ionic exchange accentuates this effect. Table 2 presents the chemical composition of the slime fraction of a n oxidized iron ore sample fractionated into three size ranges. I t is apparent that iron and alumina increased in the f iner size ranges, whereas silica decreased. From x-ray di f f rac t ion analyses, differential thermal analyses and electron microscopy, the slime fraction was shown to consist of hematite, goethite, quar tz and kaolinite. I t is interesting to note that the -1 p m fraction was particularly high in goethite and kaolinite.

Table 2. Analysis of a n Oxidized Iron Ore Slimes

................................................. ................................................. O/o Fe % SiO, % AI,03 % H,O

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Head 18.68 69.80 0.26 1.70 a t 400°C 20/5 pm 19.63 67.92 0.43

5/1 pm 24.06 64.60 0.75 -1 m 56.71 1 1.70 2.68 2.01 a t 400°C

2.74 a t 800°C ................................................. ................................................. Iwasaki, Cooke, Harraway and Choi (1962)

Attempts to compensate by the addition of large amounts of collectors result in only partial correction. Table 3 shows a n example of the flotation results on an oxidized iron ore using dodecylammonium chloride as a collector. When the same amount of amine was added to the deslimed and undeslimed samples, the collector was consumed by the slimes and the recovery of the siliceous gangue was markedly reduced. When the amount of the collector addition to a deslimed ore was reduced to one half , the solution concentration of the amine was nearly identical with the undeslimed sample, yet the floatabili ty of the siliceous gangue was not as severely affected. Apparently, slime coating on granular particles interfered with the bubble-mineral contact. I t is now fairly well established that the electrostatic interaction between slimes and mineral particles governs the slime coating, and that the nature of the slime coating may be interprcted by the heterocoagulation theory (Fuerstenau et al., 1958; Usui, 1972).

ADVANCES IN COAL AND MINERAL PROCESSING USING FLOTATION

Table 3. Ca t ion ic Silica F lo ta t ion Results on an Oxid ized I ron O r e IJsing Dodecylammonium Chloride.

...................................................................................... ...................................................................................... Amine Con-

Pct S iOz cent ra t ion A f t e r Pct Collector F lo ta t ion Pct Pct Pct Recovery in Condi t ion ing , Abstract ion Addi t ion Products Wt F e Insoluble Flotat ion m g per L of Amine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

0.15 k g / t F e conc. 20.43 59.61 10.56 (dcsl imed) S iOz f r o t h 65.35 4.27 91.43 96.52 10.0 70.7

sl imc 14.22 25.21 59.32

0.075 k g / t F e conc. 57.54 24.49 62.41 (dcsl imed) S i 0 2 f r o t h 28.81 2.25 95.30 43.33 3.5 79.5

sl imc 13.65 26.02 58.15

0.15 kg/ t F e conc. 84.93 19.01 7 1.44 S i 0 2 f r o t h 15.07 12.83 79.46 16.48 3.4 90.0

...................................................................................... ......................................................................................

Iwasaki , Cookc. H a r r a w a y a n d Choi (1962)

A n e l e g a n t m e t h o d o f v e r i f y i n g t h e heterocoagulat ion theory was reported by Usui et al. (1967) by using twin mercury drople t electrodes. By v a r y i n g t h e polarizat ion potentials of t h e mercury drople t s in potassium f luor ide solutions, they established t h e c r i t i ca l potentials f o r the i r coalescence. T h e results a r e shown in F i g u r e 6. T h e curves d r a w n through the exper imenta l ly observed c r i t i ca l potentials represent t h e potentials a t the ou te r Helmholtz plane ($ ) calculated 6 independent ly f r o m t h e d i f f e r e n t i a l capac i ty da ta .

T h e electrostatic in te rac t ion model of s l ime coa t ing o n q u a r t z by f e r r i c ox ide slimes, shown in F igure 7, may be compared wi th t h e domains of coalescence in F igure 6 . T h e applicd potentials , El and E,, may bc assumed to represent the s u r f a c e potentials ( $ o ) of q u a r t z a n d hemat i te , respectively, by

(Clo = 59.2(pHpZc - p H ) mV

T h e p H axis in thc f i g u r e was d r a w n in such a m a n n e r t h a t when t h e q u a r t z is a t its P Z C ( p H 3.7 in Fuers tenau e t al. (1958)), t h e s u r f a c e po ten t ia l of the f e r r i c ox ide slimes w i t h t h e P Z C of pH 8 is +278 mV. Conversely, when t h e f e r r i c ox ide is a t i ts PZC, the s u r f a c e potential o f t h e q u a r t z is -278 mV. T h e intersections wi th t h e c r i t i ca l potential curves in the f i r s t a n d t h e th i rd q u a d r a n t s d e f i n e t h e pH range where heterocoagulat ion c a n occur. Although f u r t h e r re f inement is necessary in view of t h e dissimilar sur faces wi th d i f f e r e n t H a m a k e r coef f ic ien ts (Usui , 1972), i t is interest ing to note t h a t t h e f e r r i c ox ide coa t ing densi ty o n q u a r t z in F i g u r e 7 wi th a m a x i m u m near p H 6 is in good agreement wi th t h e twin mercury droplc t in te rac t ion d a t a , a n d t h a t t h e sl ime coa t ing c a n occur evcn when the two minera l s a r e similarIy charged . Coagula t ion of the f e r r i c ox ide sl ime tends t o s h i f t the m a x i m u m of the a p p a r e n t coa t ing dens i ty c u r v e to near t h e P Z C of t h e slime.

T h e sizes of goe th i te s l ime part icles a t t a c h e d t o q u a r t z s u r f a c e s t h a t i n t e r f e r e wi th ca t ion ic f lo ta t ion were de te rmined under a scanning electron microscope to be essential ly less t h a n I p m a n d most a r c smaller t h a n 0.5 p m in d i sc re te part icles (Tippin , 1974). Usui a n d T a k e d a (1983) also reported t h a t hemat i te s l ime part icles

F I G U R E 6. Cr i t i ca l Potentials in the Coalescence of Mercury DropIets i n Aqueous K F Solutions. T h e p H R a n g e f o r Heterocoagulat ion of Fer r ic O x i d e Slime o n Q u a r t z is Ind ica ted ( a f t e r Usui , Yamasaki , a n d Shimoi izaka , 1967).

o f less t h a n 0.3 p m were observed o n coarse q u a r t z particles. I t fol lows then, t h a t q u a r t z part icIes wi th i ron oxide sl ime coa t ing a n d wi th n a t u r a l s tains c a n b e removed by high solid sc rubbing w i t h e i thcr s u l f u r i c ac id or caust ic soda i n v iew of t h e coa t ing behaviors shown i n F igures 6 a n d 7.

T h e principles described here were well i l lustrated in t h e roles t h a t d i f f e r e n t mineralogical components p lay in t h e f lo ta t ion separa t ion of ox id ized i ron ores ( Iwasaki


1 G Slime Per Liter 0 2 G Slime Per Liter i

FIGURE 7. Effect of Slime Concentration and pH on the Ferric Oxide Slime Coating Density on Quartz in the Absence of Collector (Fuerstenau, Gaudin and Miaw, 1958).

e t al., 1962). In the cationic flotation of siliceous gangue a t near neutral pH, clay slimes are the reagent consuming species, but if enough amine collector is added so that the solution concentration is maintained, the clay slimes do not interfere with the flotation. Conversely, goethite slimes do not consume the amine collector, but they seriously interfere with the flotation of quartz particles because the goethite slimes coat the particles and prevent the attachment of a i r bubbles to the quartz surface. In the anionic flotation of goethite, the goethite slimes are the reagent consuming species, but they do not interfere with the flotation so long as sufficient dodecylsulfate collector is added. In this case, negatively-charged quartz and particularly clay slimes coat the positively-charged goethite surface and interfere with flotation.

The optimum pH for anionic silica flotation is over I I, so both iron oxide slimes and quartz particles are negatively charged. Here, the slime interference is much less pronounced. It is interesting to note that apparently cationic silica flotation a t pH 10-10.5 along with a starch depressant is also more tolerant of slimes. Bunge et al. (1977) reported that the Sept-Iles plant operated successfully without desliming on the Knob Lake hematite ore ground to -149 p m (100 mesh).

Desliming in laboratory flotation tests is commonly carried out by repeating the agitation-settling- decantation procedure a t 10 to 20 p m quartz equivalent using caustic soda and/or sodium silicate as dispersants. In pilot and commercial plants, hydrocyclones are used for desliming a t 5 to 20 Pm. Normally, dispersants are not required. Any attempt to develop a desliming device capable of splitting a t a finer size than 10 p m is subject to the success of flotation separation in the size range of 10/1 pm. Besides slime coating, reagent consumption and heterocoagulation (or co-flocculation), particles in this size range are known to cause problems

in flotation by entrainment, slower flotation rate and more reactive surfaces (Warren, 1984).

Selective Desliming

When finely disseminated ores a re ground to liberation, conventional desliming leads to an excessive loss of iron units. Selective flocculation of iron oxides followed by the removal of dispersed siliceous slimes offers one of the most promising approaches for the treatment of finely disseminated, oxidized iron ores (Frommer, 1964; Colombo, 1986). The selective flocculation is brought about by the use of a proper combination of dispersant and a starch flocculant, and the deslimed products are fur ther upgraded by the cationic flotation of the remaining siliceous gangue. The process has been successfully applied to a Michigan hematitic ore (Villar and Dawe, 1975).

High intensity magnetic separation is also capable of functioning as a selective desliming device for oxidized iron ores, rejecting siliceous gangue particles as well as slimes, while retaining f ine iron oxide particles. The magnetic concentrates thereby obtained can then be upgraded by a combination of cationic and anionic silica flotation (Lawver et al., 1965).

To achieve the ful l benefit of selective flocculation, the ground pulp must be properly dispersed prior to the addition of a starch flocculant, but the optimum conditions are extremely sensitive to the mineralogical characteristics of the ore and the type and amounts of soluble salts in pulp solutions (Iwasaki, 1979; Paananen and Turcotte, 1980). For some iron ores, caustic soda alone is sufficient and fo r others, a combined use of caustic soda and sodium silicate works well. However, more generally, sodium tripolyphosphate in combination with caustic soda and sodium silicate may be necessary.

The alkaline earth cations in pulp solutions may be deactivated in several different ways. Sodium silicate is a nonselective dispersant, but can be used effectively in selective desliming when calcium ion concentration is low, say less than 20 ppm. At low silicate additions, calcium silicate precipitates coat the surfaces of both quartz and iron oxides, but the hydrophilic patches on the minerals are sparse and starch molecules can be adsorbed with little hindrance f rom the precipitate coating. At higher levels of sodium silicate, however, a thick coating of the precipitates on iron oxide surfaces shields starch molecules f rom adsorption (Krishnan and Iwasaki, 1982). Sodium tripolyphosphate (STPP) and sodium hexametaphosphate (SHMP) behave in a similar manner as sodium silicate below stoichiometric addition. They form calcium polyphosphate precipitates that coat both quartz and iron oxide surfaces and interfere with the starch adsorption. At higher concentrations of polyphosphates, most of the precipitates are detached through complexation and dissolution. Sufficient amounts of the calcium polyphosphate precipitates, however, remain on the iron oxide surface and interfere with selective flocculation (Heerema, et al., 1982).

Figure 8 shows the results of selective desliming tests on an equal weight mixture of goethite and quartz, plotted as grade-recovery curves. The vertical line drawn at the head grade of 28.5% iron represents no selectivity of separation, corresponding to the conventional desliming on a fully dispersed pulp. In the selective desliming, the selectivity of separation improves


F l G U R E 8. Selective Desliming Results on a n Ar t i f ic ia l Mixture of Geoth i te a n d Quartz a t pH I I ( a f te r Heercma. Lipp, and lwasaki , 1983).

100

90 - C Q)

5 80-

i. Q) > 8 70 Q)

a Q) L L

60

as the curve extcnds f a r t hc r to t hc r ight . Thc cu rvc wi th no Ca+l, bu t wi th corn s ta rch , shows high selectivi ty of separation. In thc presencc of Ca+ l wi th or wi thout starch addi t ion , the pulp was total ly flocculated a n d very l i t t le slimes were removed. T h c ef fec ts of EDTA, STPP a n d SHMP a re compared a t 150°/o of the stoichiometric addi t ion fo r M Ca+'. T h e selectivi ty of separation call be ru l ly restored by complexing the calcium ion wi th EDTA. STPP is seen to be only part ial ly effective whereas SHMP fu l l y disperses the pulp a n d is total ly ineffective. I t is ev ident t h a t selective f locculat ion is appreciably more sensit ive to the na tu r e of calcium compounds formed wi th these complexing agents than cationic silica f lotat ion.

I I I I I I I -

- -

-

- o No slarch - O Slarch added

1 6 ' ~ ~a':starch added V) I o I No complexant - z I - - I 150% EDTA 0 1 a 1 A 150% STPP

C, C I 2 I v 150%SHMP

&LALp'L ' 25 30 34 38 42 4 5 50

In the presence of calcium ion i n pulp solutions in excess of , say 20 ppm, therefore, the selective f locculat ion of iron orcs can be implemented sat isfactori ly wi th the use of a strong complexing agent, such as EDTA and NTA (Learmont et al., 1984). With some iron ores containing gypsum, the calcium ion concentrat ion may reach several hundred ppm. Thei r dosage ra tes would then become prohibit ively high a n d also the complexed calcium ion would accumula te in recycled water . Fur thermore , the use of these complexing agents may be undesirable f o r their toxicological a n d ecological properties.

Percent Fe in S a n d s

Clay minerals , part icularly montmoril lonite, wi th large cation exchange propert ies may be added to remove calcium ion in solution, a l though their presence appears to make selective Flocculation more sensit ive to opera t ing condit ions. In the presence of calcium ion, montmoril lonite causes ind iscr imina te dispersion of both hemati te a n d quar tz , and hemati te part icles d o not f locculate upon addi t ion of starch. Heterocoagulat ion of hemati te a n d montmoril lonite may be intensif ied in the presence of calcium ion. I t is interest ing to note in F igure 9 t ha t ul trasonic t rea tment of the pulp markedly improves the f locculat ion of hemati te when s ta rch is added to the pulp before the t rea tment , but no beneficial e f f ec t may be observed when starch is added after the ul trasonic treatment. Apparent ly , the ultrasonic t rea tment leads to the replacement of montmoril lonite part icles by s ta rch molecules a t the hemat i te surface. T h e addi t ion of starch a f t e r the

Ultrasonic Time, seconds

a 0

V 30

v 120

Percent Fe in Sands F I G U R E 9. Selective Floccualt ion of a Hcmati te-Quartz- Mixture in the Presence of 1000 mg/L Montmoril lonitc a n d 10 mg/L Calc ium ion as a Funct ion of l i l t rasonic T rea tmen t Time, 30 mg/L Starch (added before sonication), pH 11 (Arol a n d Iwasaki , 1987).

ul trasonic t rea tment does not a id f locculat ion since montmoril lonite part icles re-deposit on the hcmati te su r f ace a n d prevents the s ta rch adsorption (Arol a n d lwasaki , 1987).

Chemical precipitat ion per se is iner fec t ive d u e to su r f ace precipitat ion as well as heterocoagulat ion of the precipitates wi th qua r t z part icles (Heerema a n d lwasaki , 1980; Kr i shnan a n d Iwasaki , 1986). However, chemical precipitat ion of calcium ion a s a carbonate salt fol lowed by mechanical cleaning of the sur face precipitates through ultrasonic t rea tment shows considerable promise in controll ing calcium ion in pulp solutions (Manukonda a n d Iwasaki , 1987). A qua r t z suspension in the presence of as h igh as 400 ppm ~ a + ~ dispersed fu l l y wi th about a half of the stoichiometric amount of bicarbonate a f t e r ul trasonic t rea tment . T h e dissolution of qua r t z might have been accelerated a t the prevail ing pH of 11 by the ultrasonic t rea tment , thereby precipitat ing the remaining Ca+2 wi th the si l icate ion. Scanning electron microscope observations of ul trasonically dispersed qua r t z samples show no precipitate coating. Selective f locculat ion tests, conducted on a hemati te-quartz mixture in the presenee of 400 ppm Ca+l, a n d ultrasonic t rea tment fol lowing a n addi t ion of the stoichiometric amount of bicarbonate produced sa t i s fac tory separation.

E F F E C T OF PARTICLE S IZE

As mentioned earl ier , on well-sized, s l ime-free particles, the selectivi ty of f lotat ion separation f rom ar t i f ic ia l mixtures may be related to t he d i f f e r ence in their PZC's. However, the role t ha t part icle size plays in t he f lo ta t ion separation of oxide minerals wi th paraf f in-cha in collectors is of ten overlooked. In the processing of magnetic taconite, f i n e siliceous f rac t ions a r e well l iberated a n d removed in the magnetic separation step, while the size f rac t ions coarser t han 44 pm (325 mesh) become progressively higher in the SiO, contents, a n d more than half of the silica is i n the +44


Table 4. Screen Analyses of a Magnetic Taconite Concentrate and Its Flotation Concentrate. The Effect of Particle Size on Flotation Removal of Silica is Included.

........................................................................

Screen Size, Magnetic Concentrate Flotation Concentrate* SiO, Floated Mesh % wt % Fe % SiO, % wt % Fe % SiO, %

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 0.70 35.94 40.08 0.60 36.27 40.48 13

Composite 100.00 67.28 4.91 79.56* 68.34 3.97 36 5------P- ..............................................................

* Cleaner concentrate a t point A in Figure 10.

pm (+ 325 mesh) fractions. Table 4 shows a typical example of the screen analysis on a magnetic taconite concentrate. Therefore, the cationic silica flotation of magnetic taconite concentrates involves the flotation of coarse and locked siliceous gangue particles f rom the ful l size range of magnetite particles.

Figure LO shows the cationic silica flotation results on the magnetic concentrate. With increasing addition of the collector at the natural pH of 8.5, the iron recovery drops precipitously a t about 3.6% SiO, in the flotation concentrate. The screen analysis of the cleaner concentrate in Table 4 shows that the SiO, contents in the size fraction coarser than 44 p m (325 mesh)

0 Nalural pH (8.5)

0 +325M Csnc B -325M Conc. MI. OH

t 5 W M Conc B -500M Conc, nat pH C,

5 4 1 I I I I I I I I

I I 67 60 69 70 71

FIGURE 10. Effects of pH, Caustic Scrubbing and Coarse-Fine Separation on the Flotation of a Magnetic Taconite Concentrate Using a n Ether amine. The Results of Screnning Out the Coarse Sizes f rom the Magnetic Concentrate (Table 4) a re Included fo r Comparison.

remained quite high and the SiO, recoveries in the cationic silica flotation in the +44 p m (+325 mesh) fractions were notably poor. This is because the coarse gangue particles have the same floatabilities as the f ine magnetite particles, and fur ther addition of the collector floats both together indiscriminately. In fact , the flotation results are seen to be less selective than simply screening out the coarse fractions. This behavior becomes more critical when the siliceous gangue is in the form of ferromagnesian silicates with their PZC's between quar tz and magnetite as shown in Figure 1.

Frothers can reduce the amount of cationic collectors approximately weight-for-weight, but with very little ef fect on selectivity. Kerosene and fuel oil a re used effectively as a n extender in the anionic flotation of oxidized iron and phosphate ores, but their use along with cationic collectors will not float coarse siliceous gangue in magnetic concentrates. The effect of pH, as seen in Figure 10, is relatively minor though somewhat higher selectivity was obtained a t pH 10 than a t 8.5 and 5.5. High solid scrubbing with caustic soda at pH 12.3 followed by desliming and flotation a t pH 10 appeared to improve the selectivity of flotation separation (Anon, 1977).

To search fo r more selective collectors, both company and university laboratories have carried out a large number of flotation tests with d i f ferent types of amines (Bleier et al., 1976; Cronberg and Lentz, 1968; Freyberger et al., 1981; Jacobs, et al., 1978; Smith, 1973). Certain amines are somewhat more selective than others, and in commercial plants, ether amines and diamines wi th the following formulas

Ether Primarv Amine Ether Diamine

R-0-(CH,), -N A R-O-(CH2)3-HN-(CH2)3 -N / H

'H 'H

where R is C,/Clo where R is a C,, branched hydrocarbon ~n a hydrocarbon chain ratio of 60/40


are used for ease of feeding, metallurgical efficiency and cost. The diamines are used as partially neutralized with acetic acid, or unneutralized. The optimum pH fo r the cationic silica flotation of oxidized iron ores is in the range of 10 to 11, while the flotation of magnetic taconite concentrates is carried out a t the natural pH of near 8.5.

Another approach to this problem would be to search for more selective depressants for magnetite. Starches and starch derivatives have been used successfully in the flotation of oxidized iron ores. In Figure Il(a), the ef fect of various starch depressants is shown as grade-recovery curves, replotted f rom Chang, Cooke and Huch (1953). The figure shows that certain starches are morc effective than othcrs. When the starch depressants wcre tested on magnetic taconite concentrates, however, they had virtually no effect on the selectivity of separation as shown in Figure I l(b).

Derrick sandwich deck screens are available to 37 pm (400 mesh) and they can operate effectively on a commercial scale (Derrick et a]., 1989). Because of the heavy media ef fect on the vibrating screen, the siliceous gangue particles tend to float and report selectively to the oversize. I t becomes of interest then to study the effects of screening a flotation feed into coarse and f ine fractions, and carrying out the flotation tests on the two size fractions separately as described earlier.

In the selective desliming-silica flotation applied to oxidized taconites, the concentrate grade could be improved by adding more amine collector to ensure the flotation of coarse gangue particles. The increased addition of the collector causes the flotation of f ine iron oxides leading to a loss of iron units. Wet high-intensity magnetic separation of the cationic silica flotation froths was reported to recover the f ine iron oxides, which was recycled to the overall flotation circuit,

(a) Hematite Ore

0 None 0 Amylopectin

Amylose Corn

A BritishGum

50 55 60 65 % Fe in Concentrate

- (b) Magnetite Conc - 0 None 0 Cationic Gum - -

Anionic Gum A BritishGum A Dextrin

- -

I I I 65' 66 67 68

% Fe in Concentrate

FIGURE 11. Effect of Various Starches and Their Derivatives on Cationic Flotation of (a) Hematite Ore (af ter Chang, Cooke and Huch, 1953), and (b) Magnetite Concentrate (Iwasaki, Carlson and Parmerter, 1969).

The size dependence of their floatabilities may be utilized in the upgrading of a magnetic taconite concentrate. When a concentrate is sized into coarse and f ine fractions with a laboratory sieve and they are processed separately, the combined concentrate f rom the coarse and f ine circuits is markedly higher in grade than when the unsized magnetic taconitc concentrate is floated directly (Figure 10). Conventionally, hydrocyclones are used in commercial operations, but cyclones classify on the basis of particle size and specific gravity, and closing of a grinding circuit with a cyclone forces coarse siliceous gangue into the overflow fraction. The replacement of a cyclone by a Rapif ine screen reportedly minimized coarse silica and resulted in a better f lotation response for the Knob Lake hematite ore (Major-Marothy, 1972).

thereby increasing the overall iron recovery by 2 percent with no adverse effect on the concentrate grade (Hopstock, 1979). The effectiveness of f ine screening the f ro th product may also merit attention.

DUPLEX FLOTATION

The size dependence of the floatabilities of iron oxides and siliceous gangue may be utilized in the combination or duplex process applied to low-grade oxidized iron ores. Fat ty acid flotation of iron oxides following desliming results in an appreciable contamination of the concentrate by fine-sized siliceous gangue due presumably to the accidental activation by calcium ion. This is exactly opposite in behavior to the cationic flotation of siliceous gangue f rom magnetic


concentrates in which silica f ro ths were contaminated with f ine magnetite. Therefore. if the fa t ty acid coating on iron oxides can be removed, the f ine siliceous gangue can be removed by the cationic silica flotation by taking advantage of the differences in the floatabilities between coarse and f ine particles in both steps. This combination process is successfully applied to the beneficiation of Florida phosphate ores in which the f a t ty acid coating is removed by acid scrubbing. However, acid scrubbing is totally ineffective o n iron ores.

Two methods have been proposed fo r removal of the hydrophobic coating of f a t ty acids f rom iron oxide concentrates. In one method, activated carbon with a large specific surface of the order of 600 mZ per g was used to remove the adsorbed collector by distribution over its enormous surface area. An oxidized iron ore, analyzing 37% iron, was ground to -149 p m (-100 mesh), deslimed and upgraded by flotation using 0.25 kg/ton fa t ty acid. The concentrate analyzed 44% iron, which was heavily contaminated with fine siliceous gangue. The concentrate coated with fa t ty acid was then conditioned with 1.5 kg/ton activated carbon. The activated carbon was then removed by flotation without any addition of collector, followed by the cationic flotation of siliceous gangue with 0.3 kg/ton British gum as a depressant fo r iron oxides. The f inal concentrate analyzed 62.5% iron with 69% iron recovery (Iwasaki, Zetterstrom and Kalar, 1967).

The second method involves the use of ozone fo r decomposing the f a t ty acid coating. Ozone can oxidize organic compounds rapidly without leaving potentially hazardous by-products. A hematite ore, analyzing 36% iron, was processed by the same flowsheet as shown in Figure 12. T h e iron concentrate with f a t ty acid coating

Hematite Ore (36Y Fe) t

Grind to -100 mesh + Desliming (20prn)

I

Slime Sands

(H47:; 3 0.25 kglt Soda Ash 0.25 kgn Fany Acld

Fe Flotation

r l Fe Conc Tails

6Ph Wt 14% wt (45% Fe) (13% ~ e )

0.25 kgR o3 1 Conditioning

0.25 k /t Dextrl 0.3 kglt E t ~ e r Arnmn-l

Si02 Mids ~ e ' c o n c Froth

( ?2Gt) FIGUKE 12. Duplex Flotation of a Hematite Ore Involving Fat ty Acid Flotation of Hematite, Ozone Treatment and Cationic Silica Flotation (af ter Iwasaki and Malicsi, 1985).

was treated by bubbling ozone. The hydrophobic coating was completely destroyed with 0.25 kg/ton ozone. T h e amine flotation of siliceous gangue f rom a n ozonated pulp produced a f ina l concentrate analyzing 65% iron a t a n overall iron recovery of 70%. This duplex flotation process appears to be capable of producing a concentrate with similar grade and recovery at much coarser mesh- of-grind than the conventional o r selective desliming- cationic silica flotation process.

SUMMARY

North American iron ores are low in grade and finely disseminated, but are relatively free f rom such impurities as sulfur, phosphorus and titania. In order to remain competitive with high-grade overseas sources, the investigations on low-cost beneficiation methods that insure high quality and uniform quality products must be continued in order to prepare for ever tightening specifications on ironmaking raw materials for blast furnaces and also fo r the fu ture iron- and steelmaking processes that can directly utilize the finely ground concentrates.

In this article, a major emphasis was placed on the problem areas of interest a t present in the cationic silica flotation. Some of the seemingly paradoxical behaviors may be related to the effects of surface contamination, inherently caused by natural processes or by unintentional pretreatments, as well as to the effects of minor impurities or lattice defects as reported on other oxide minerals (Aplan and Spearin, 1980; Balachandran et al., 1987; Mular, 1965; Roy and Fuerstenau, 1972). The development of more specific collectors or depressants becomes of interest in improving the selectivity of separation. Perhaps, f lotation reagents with chelating groups may provide a n approach to this problem (Fuerstenau et a]., 1967; Marabini e t al., 1983; Pradip, 1988). although the effect of residual reagents on the environment should be carefully assessed.

Besides silica, many other minor impurities are becoming of concern for improving the productivity and product quality in ironmaking and steelmaking. The removal of alkali metals, sulfur and phosphorus f rom iron ores, that are often directly amenable to conventional flotation, has been reviewed recently (Nummela and Iwasaki, 1983). The attention should also be directed towards removing more refractory impurities which are either intimately associated or chemically combined, such as arsenic, zinc, tin, nickel and titania.

While the surface chemistry plays a fundamental role in the selection of a proper chemical combination, attention should also be paid to the engineering aspects of the mineral-reagent as well as mineral-bubble contacts. T h e importance of conditioning has been well recognized in the anionic flotation of iron oxides (Arbiter and Williams, 1983; Chi and Young, 1962; Li, Livingston and Lemke, 1960). In the cationic silica flotation, capillary condensation of undissociated amines may play a similar role in alkaline pulps (Takeda and Usui, 1987). The effect of flotation machine design and operation has not been fully explored and warrants both theoretical as well as developmental research (Beebe, 1965). Column flotation with less vigorous agitation than mechanical cells and the ef fect of bubble size distribution on the selectivity of separation merit attention (Szatkowski and Freyberger, 1988).


REFERENCES

Anon, "Cationic Silica Flotation," 1977, Mining Magazine, Jan., pp. 25-31.

Aplan, F.F., and Fuerstenau, D.W., 1962, "Principles of Nonmetallic Mineral Flotation," Froth Flotation, Fuerstenau, D.W., ed., AIME, New York, pp. 170-214.

Aplan, F.F., and Spearin, E.Y., 1980, "Flotation of the p- Type Semiconductor NiO as a Function of Point Defects," Colloids and Surfaces, Vol. 1, pp. 361-371.

Arbiter, N., and Williams, E.K.C., 1980, "Conditioning in Oleic Acid Flotation," Fine Particles Processing, Somasundaran, P., ed., AIME, New York, pp. 802-831.

Balachandran, S.B., Simkovich, G., and Aplan, F.F., 1987, "The Influence of Point Defect on the Floatability of Cassiterite," International Journal of Mineral Processing, Vol. 21, pp. 157-184.

Beebe. R.R., 1965, "New Directions in Iron Ore Flotation," Skillings' Mining Review, Vol. 54, No. 23, June 5, PP. 1, 6-7, 15.

Bleier, A., Goddard, E.D., and Kulkarni, R.D., 1976, "The Structural Effects of Amine Collectors on the Flotation of Quartz," Flotation, A.M. Gaudin Memorial Volume, Fuerstenau, M.C., ed., AIME, New York, pp. 117-147.

Bunge, J.H., Morrow, J.B., Trainor, L.W. and Dicks, M.L., 1977, "Development and Realizations in the Flotation of Iron Ores in North America", Proceedings. XI1 International Mineral Processing Congress, Sao Paulo, Brazil. 74 pp.

Chang, C.S., Cooke, S.R.B., and Huch, R.O., 1953, "Starches and Starch Products as Depressants in Amine Flotation of Iron Ore, Trans. AIME, Vol. 196, pp. 1282- 1286.

Chi, J.W.H., and Young, E.F., 1962-63, "Kinetics of Flotation Conditioning of a Hematite Ore," Transactions, Insti tution of Mining and Metallurgy, London, Vol. 72, p p 169-190.

Clark, S.W., and Cookc, S.R.B., 1968, "Adsorption of Calcium, Magnesium, and Sodium Ion by Quartz," Trans. SME-AIME, Vol. 241, pp. 334-341.

Coburn, J.L., 1985, "Froth Selectivitv in the Flotation of Ouartz f rom Finelv Ground Oxidized lron Ores Using Primarv Branched Chain Ether Amines," PhD Thesis, Michigan Technological University, 153 pp.

Colombo. A.F., 1986, "Concentration of lron Oxidcs by Selective Flocculation-Flotation," Advances in Mineral process in^, Somasundaran, P., ed., SME, Littleton, Colorado, pp. 695-7 13.

Cooke, S.R.B., Iwasaki, I., and Choi, H.S., 1960, "Effect of Temperature on Soap Flotation of lron Ore," Trans SME-AIME, Vol. 217, pp. 491-498.

Cronberg, A.D., and Lentz, T.H., 1968, "Beneficiation of Magnetic Iron Concentrates by Cationic Silica Flotation with Amine Acetates," SME-AIME P r e ~ r i n t , 68-B-340, New York, I I pp.

Derrick, H.W., Wennen, J.E. and Nordstrom, W.J., 1989, "Applications of Fine Screens in Taconite Ore Concentration," Proceedings of the 62nd Annual Meeting of the Minnesota Section AIME and 50th Annual Mining Symposium, University of Minnesota-Duluth, 23 pp.

DeVaney, F.D., 1943, "Beneficiation of Magnetite Concentrates by Flotation," U.S. Patent 2,388,471, Nov. 3.

Freyberger, W.L., O'Brien, B.M., and Keck, J.W., 1981, "Quartz Flotation with Branched Chain Ether Amines," I n t e r f a c i a l g i n Yarar, B., and Spottiswood, D.S., eds., Engineering Foundation, New York, pp. 255-270.

Frommer, D.W., 1964, "USBM Tests on Selective Iron Ore Flotation Point Way to Higher Recoveries," Mining -, Vol. 16, pp. 67-80.

Fuerstenau, D.W., 1970, "Interfacial Processes in Minerallwater Systems," Pure and A ~ p l i e d Chemistrv, Vol. 24, pp. 135-164.

Fuerstenau, D.W., Gaudin , A.M., and Miaw, H.L., 1958, "Iron Oxide Slime Coatings in Flotation," Trans. AIME, Vol. 2 1 1, pp. 792-795.

Fuerstenau, D.W., and Fuerstenau, M.C., 1982, "The Flotation of Oxide and Silicate Minerals," P r i n c i ~ l e s of Flotation, King, R.P., ed., South African Institute of Mining and Metallurgy, Johannesburg, pp. 109-158.

Fuerstenau, D.W., a n d Raghavan, S., 1978, "The Surface and Crystal Chemistry of Silicate Minerals and Their Flotation Behavior, Freiberger Forschungshefte, A593, pp. 75-109.

Fuerstenau, M.C., Miller, J.D., and Gutierrez, G., 1967, "Selective Flotation of Iron Oxide," Trans. SME-AIME, Vol. 238, pp. 200-203.

Fuerstenau, M.C., and Palmer, B.R., 1976, "Anionic Flotation of Oxides and Silicates," Flotation. A.M. Gaudin Memorial Volume, Fuerstenau, M.C., ed., AIME, New York, pp. 148-196.

Gaudin , A.M., 1957, "Flotation," McGraw-Hill, New York, 573 p.

Heerema, R.H. and Iwasaki, I., 1980, "Chemical Precipitation of Alkaline Ear th Cations and Its Effec t on the Flotation and Flocculation of Quartz," Trans SME- w, Vol. 268, pp. 1510-1516.

Heerema, R.H., Lipp, R.J., and Iwasaki, I., 1982, "Complexation of Calcium Ion in Selective Flocculatio~n of Iron Ores," Trans SME-AIME, Vol. 272, pp. 1879-1884.

Hogg, R., Healy, T.W., and Fuerstcnau, D.W., 1966, "Mutual Coagulation of Colloidal Dispersions," Transactions of the Faradav Society, Vol. 62. pp. 1638- 1651.

Hopstock, D.M., 1979, "West High Intensity Magnetic Beneficiation of Oxidized Taconites." USBM R e ~ o r t of Investigations, RI 8363, 13 pp.

Houot, R., 1983, "Beneficiation of Iron Ore by Flotation," International Journal of Mineral Processinq, Vol. 10, pp. 183-204.


Hruz, F.M, 1954, "Amine Flotation of Quartz as A ~ v l i e d to Iron Ore Beneficiation," M.S. Thesis, University of Minnesota, Minneapolis, Minnesota, 61 pp.

Iwasaki, I., 1979, "Selective Flocculation of Fine-Grained Iron-Bearing Materials --- Principles," Beneficiation of Mineral Fines. Problems and Research Needs, Somosundaran, P., and Arbiter, N., eds., AIME, New York, pp. 257-262.

Iwasaki, I., 1983, "Iron Ore Flotation, Theory and Practice," Gaudin Award Lecture, Mining Engineering, Vol. 35, pp. 622-632.

Iwasaki, I., Carlson, W.J., and Parmerter, S.M., 1969, "The Use of Starches and Starch Derivatives in Iron Ore Beneficiation," Trans. SME-AIME, Vol. 244, pp. 88-98.

Iwasaki, I., Cooke, S.R.B., and Choi, H.S., 1961, "Flotation of Cummingtonite," Trans. SME-AIME, Vol. 220, pp. 394- 395.

Iwasaki, I., Cooke, S.R.B., and Colombo, A.F., 1960, "Flotation Characteristics of Goethite," USBM Renort of Investigations, RI5593, 25 pp.

Iwasaki, I., Cooke, S.R.B., Harraway, D.H., and Choi, H.S., 1962, "Iron Wash Ore Slimes --- Some Mineralogical and Flotation Characteristics," Trans. SME-AIME, Vol. 223, pp. 97-108.

Iwasaki, I., Cooke, S.R.B., and Kim, Y.S., 1962, "Some Surface Properties and Flotation Characteristics of Magnetite," T-E, Vol. 223, pp. 113-120.

Iwasaki, I., and Malicsi, A.S., 1985, "Use of Ozone in Iron Ore Flotation," Minerals and Metallurgical Processing, Vol. 2, pp. 104-105.

Iwasaki, I., Zetterstrom, J.D., and Kalar, E.M., 1967, "Removal of Fat ty Acid Coatings from Iron Oxide Surfaces and its Effect on the Duplex Flotation Process and on Pelletizing," Trans. SME-AIME, Vol. 238, pp. 304- 3 12.

Jacobs, W., Papacek, H.G., and Brennecke, K., 1978, "Inverse Flotation of Iron Ores," Skillinas' Mining Review, Vol. 67, No. 31, August 5, pp. 10-11, 20-24.

Johnson, E.B., and Bjorne, C., 1964, "Milling Practice a t the Republic ~ i n e , Republic, Mich.," in Millinp Methods inthe Arbiter, N., ed., Gordon & Breach, New York, pp. 161-176.

Krishnan, S.V., and Iwasaki, I., 1982, "Sodium Silicate as a Dispersant in the Selective Flocculation of Iron Ores," Trans. SME-AIME, Vol. 272, pp. 1984-1988.

Krishnan, S.V., and Iwasaki, I., 1986, "Heterocoagulation versus Surface Precipitation in Quartz - Mg(OH)2 System," n, Vol. 20, pp. 1224-1229.

Lawver, J.E., Beebe, R.R., and Hays, R.M., 1964, "New Methods fo r Beneficiating Semitaconites," Minina Congress Journal, Vol. 51, April, pp. 69-74.

Learmont, M.E., Lipp, R.J., and Iwasaki, I., 1984, "Complexation of Calcium Ion with NTA in the Selective Flocculation of Iron Ores," Trans. SME-AIME, Vol. 276,

pp. 1817-1818.

Li, K., Livingston, R.W., and Lemke, L.K., 1960-61, "Flotation Conditioning of Iron Ore with Petroleum Sulfonate," Transactions, Institution of Mining and Metallurgy, London, Vol. 70, pp. 19-31.

Major-Marothy, G., 1972, "Silica Flotation and the Use of Fine Screening in Iron Ore . Beneficiation," Canadian Mining J o u r n d , Vol. 93, No. 6, pp. 31-34,

Manukonda, V.R., and Iwasaki, I., 1987, "Control of Calcium Ion via Chemical Precipitation - Ultrasonic Treatment in Selective Flocculation," Mineral and Metallurgical Processing, Vol. 4, pp. 216-222.

Marabini, A.M., Barbaro, M., and Ciriachi, M., 1983, "Calculation Method for Selection of Complexing Collectors," Transactions, Institution of Mining and Metallurgy, Vol. 92, pp. C20-C26.

Mular, A.L., 1965, "Effect of Impurities on the Flotation Behavior of Zinc Oxide," Trans. SME-AIME, Vol. 232, pp. 204-2 1 1.

Nummela, W., and Iwasaki, I., 1986, "Iron Ore Flotation," -inMineral Somasundaran, P., ed., SME, Littleton, Colorado, pp. 308-342.

Onoda, G.Y., and Fuerstenau, D.W., 1964, "Amine Flotation of Quartz in the Presence of Inorganic Electrolytes," Congress, New York, N.Y., Vol. I, pp. 301-306.

Paananen, A.D., and Turcotte, W.A., 1980, "Factors Influencing Selective Flocculation-Desliming Practice a t the Tilden Mine," Trans. SME-AIME, Vol. 268, pp. 1244- 1247.

Parks, G.A., 1965, "The Isoelectric Points of Solid Oxides, Solid Hydroxides, and Aqueous Hydroxo Complex Systems," Chemical Reviews, Vol. 65, pp. 177-198.

Pradip, 1988, "Application of Chelating Agents in Mineral Processing," Minerals and Metallureical Processing, Vol. 5, pp. 80-89.

Roy, R., and Fuerstenau, D.W., 1972, "The Effect of Doping on the Point of Zero Charge of Alumina," Surface Science, Vol. 30, pp. 487-490.

Scott, D.W., Richardson, A.C., and Arbiter, N., 1945, "Amine Flotation of Gangue from Magnetite Concentrates," AIME Mining Technoloav, T.P. 1902, 10 PP.

Somasundaran, P., and Lee, L.T., 1982, "Polymer- Surfactant Interactions in the Flotation of Quartz and Hematite," Prenrint. XIV International Mineral Processing Congress, Toronto, Canada, IV-9.1-9.17.

Smith, R.W., 1973, "Effect of Amine Structure in Cationic Flotation of Quartz," Trans. SME-AIME, Vol. 254, pp. 353-357.

Smith, R.W., and Akhtar, S., 1976, "Cationic Flotation of Oxides and Silicates," Volume, Fuerstenau, M.C., ed., AIME, New York, pp. 87- 116.


Szatkowski, M., and Freybergcr, W.L.. 1988, "The Effec t of Bubble Sizc Distribution on Selectivity of Iron Ore Flotation," v g , Vol. 23, pp. 213-227.

Takeda, S., and Usui, S., 1987, "Adsorption of Dodecylammonium Ion on Quartz in Relation to its Flotation," Colloids and Surfaces, Vol. 23. pp. 15-28.

Takeda, S., and Usui, S., 1988, "Cationic Flotation of Quar tz from an Artif icial Mixture with Hematite Using Hexylamine," Colloids and Surfaces, Vol. 29, pp. 221-232.

Takeda, S., Usui, S., and Matsuoka, I., 1987, "Flotation Separation of Coarse Quartz f rom a Mixture of Hematite Fines in Alkalinc Media in Relation to the Adsorption of Dodecylammonium Acetatc," N i ~ p o n Kogvo Kaishi, Vol. 103, pp. 859-864.

Tippin, R.B., 1974, "lnflucnce of Thermal Pretreatment on Flotation Characteristics of Iron Ore," PhD Thesis, University of Minnesota, Minneapolis, Minnesota, 269 pp.

Usui, S., 1972, "Heterocoagulation," Progress in Surface and Membrane Science. Vol. 5, Danielli, J.F., Rosenberg, M.D., and Cadenhead, D.A., cds., Academic Press, New York, pp. 223-266.

Usui, S., and Takeda, S., 1983, " Effect of Particle Size on Reverse FIotation of Quartz f rom Hematite Using Dodecyiarnine in Acidic Media," Nippon Kogvo Kaishi, Vol. 99. pp. 915-920.

Usui, S., and Takeda, S., 1983a, "Efl'cct of Particle Size on Reverse FIotation of Quartz f rom Hematite Using Dodecylamine in Alkaline Media," NioDon Kogyo Kaishi, Vol. 99, pp. 995-998.

Usui, S., Yamasaki, T. , and Shimoiizaka, J . , 1967, "Coalescencc of Mercury Droplets in Aqueous Solutions, with Special Reference to the Examination of Double- Layer Interaction," Journal of Physical Chemistry, Vol. 71, pp. 3195-3202.

Villar, J.W. and Dawe, G.A., 1975, "The Tilden Mine-A New Processing Technique f o r Iron Orc," Mining Congress Journal, Vol. 61, pp. 40-48.

Warren, L.J., 1984, "Ultrafine Particles in Flotation," Principles of Mineral Flotation. The Wark Svmoosium, Jones, M.H., and Woodcock, J.T., eds, Australasian Insti tute of Mining and Metallurgy, Victoria, Australia, p p 185-213.

Documents

Bridging Theory And Practice In Iron Ore Flotationyunus.hacettepe.edu.tr/~obicak/MAD454/Iron Ore_Iwasaki.pdf · 2017-03-06 · BRIDGING THEORY AND PRACTICE IN IRON ORE FLOTATION 179