Reprinted from: XVth International Mineral Processing ...

Reprinted from: XVth International MineralProcessing Congress, Proceedings,Vol. IV, 1986, pp. 129-140.

m 4. - "Surface precipitation of surfactants aM inorganies on mineral solids aM Its

roie in 8d3Orption aM notationR K.P. ANANTHAPADMANABHAN aM

P. SOMASUNDARAN

Q. from Pro D.W. FUERSTENAU (University of Callfomia - USA)

Fil"St of all. I would like point out that at the Cagliari IMPC. discussed and showed

conditions of chemisorption. surface reaction arid bulk precipitation In the

chalcocite-dithiophosphate system. Secondly. in discussing, surface solubility

products or surface precipitation. one should make his analysis in terms of the

constancy of electrochemical potentials in.~he system.

r

My question is that surface precipitation should take place as a phase change. On

looking at your data on flotation recovery versus collector concentration, one

observes that a 5 or 10 Cold addition of collector is required for flotation from the

initiation of flotation to Cull flotation. This does not reflect surface precipitation,where one would expect a more abrupt change. How do you explain this?

)

A. from Pro P. SOMASUNDARAN\

Surface precipitation is as different from surface reactions as bulk precipitationis different from bulk reactions. Such reactions do not necessarily lead to

precipitation which is governed by additional thermodynamic criteria as representedby solubility products. indeed. electrochemical potentials should be considered while

examining surface precipitation. The existence of electrochemical potential is

precisely why surface precipitation can often occur in certain systems even in the

absence of bulk. precipitation.

The absence of a vertical rise on the flotation curve with increase in surfactant

concentration can easily be accounted for by considering the depletion of surfactant

from bulk solution under conditions of incipient surface precipitation and incipient

notation. Note that flotation recoveries are reported in literature as a function of

total added surfactant and not as a function of equilibrium or residual surfactant

concentration.

Thus, just above the concentration corresponding to surface precipitation (and

incipient flotation) there will not be enough surfactant to produce the precipitate

coating of the particle required for complete flotation. Under such sparse coverageconditions, only partial flotation can be expected. As the surfactant dosage is

129

../.

increased there will be more surfactant available in the system to more fully coat the

particles and produce increased flotation.

Under sparse coverage conditions. sut"tace heterogeneity can be expected to cause

selective flotation of the fraction of pat"ticles that at"e most active. Also because of

such heterogeneity. surface precipitation can begin prefet"entially on cet"tain active

sites at a given concentt"ation and. low enef'gy sites can require higher concentl'ations

even tot" the onset of surlace precipitation.

Yet another factor which can influence the shape of the flotation curve is the

kinetics. Flotation conditioning is done for a relatively short time compared to that

required Cor equilibrium adsorption/surface precipitation. This can lead to lower

adsorption than the expected equilibrium value.

In certain systems, precipitation may occur both on the mineral surface and in the

bulk solution. In this case, a significant portion of the surfactant can become

uQ.vailable for flotation. . .

In such systems, flotation recovery should be correlated with the amount of the

precipitate on the mineral surface.

Results for the flotation of cuprite-salicylaldoxime and chrysocoUa/LIX systemsdiscussed in detail elsewhere (II. 13) clearly demonstrate the important role of

surface Ys. bulk precipitation and how the surface precipitation correlates with

flotation.

Q. from J.L. CECILE (BRGM, Orleans - Prance)

Have you tried to correlate the residual concentrations of collector anion and o(

cation observed with concentrations calculated with thermodynamic datas ?

In (act. except (or systems carbonates/fatty acids for which there is a good

correlation (see the results obtained by PREDALl and CASES). there is in

general a very poor agreement between calculated and observed values (see for

example the system malachite-salicylaldoxime).

This means that a new phase is formed, the properties of which are very different

of those of a precipitate cation-collector anion.

130

A. (rom Pro P. SOMASUNDARAN

.. We have not correlated the residual concentrations of cation and C!ollector ion

with their values calculated using thermodynamic data. While this will be useful,

it will be equally important to monitor also the changes in the electrokinetic

potential a the solid-liquid interface and its influence on the distribution of

various ions in the interfacial region. These aspects are C!urrently being studied.

b. In the case of' carbonate minerals such as malachite. it is extremely important to

consider the role of' atmospheric CO2 in determining solubility characteristics.

Our calculations clearly show that in the case of carbonate minerals. the

solubility and consequently the stability characteristics of various solids can be

totally different depending upon whether or not the system is open to the

atmosphere. This may be one of the reasons why thermodynamic calculations havenot predicted the system behavior accurately in the case of' malachite-

salicylaldoxime.

Q. from Pro J. LASKOWSKI

( would appreciate it very much if professor SOMASUNDARAN coud clarify what

is the difference between his prei!ipitation concept and old du RIETZ precipitation

theories.

A. from PI'. P. SOMASUNDARAN

Our approach is totally different from that of du Rietz. We have shown that it is

possible to have precipitation in the interfacial region even before conditions for

precipitation are met in the bulk solution. In the absence of electrostatic or other

adsorption forces, we believe that du Rietz's approach is valid and. both the surfaceand bulk precipitation can take place under the same solution conditions.

The situation is, however, very different when the particles are charged in

solution.

Surface precipitation of Ca(OH)2 on silica and the formation of hemi-micelles inthe case of sulfonate-alumina system are among the examples in which surface

charge leads to precipitation in the interfacial region under conditions when reactionsleading to precipitation in the bulk solution are not expected theoretically.

131

Q. fro. Pr. J.D. MILLER (Unl.enity of Utah - USA)

While it Is true that the term chemisorption hu been used rather loosely In past

literature reprdlnr coUector adsorption phenomen.. it has been known for some time

from la spectroscopy that the nature of the cbem*rtled species varies with the

extent of adsorption.

For example, u early u 1963 it wu clear for semi-soluble salt minerals that

oleate does truly chemisorb at the surface as evidence by a distinct shift In the

carbonyl stretchinr frequency<I,2,3,4).

Similar characteristic features have been observed for sulflde(S) and other

nonsulflde(6,7) notation systems. Most investigators reeornize that adsorption in

these systems may include both chemlsorbed species as well u a surface precipitate

similar in character to the bulk phase precipitated eoUeetor salt(8,9,lO).

Recent studies support this anaiysis(II). As shown in nc. I. it i. evident that the

adsorption density of oleate by fluorite does exhibit a marked increase after I mr/l(3.5 ~ 10-6 M) as a result of the surfaee precipitation.

However, complete flotation of fluorite was achieved at oleate concentrations far

below this critical oleate concentration. For example, 100 ~ fluorite recovery can be

obtained at an oleate concentration of 0.3 mg/l (1 x 10-6 ~) ; similar to that obtained

by SORENSON(12) who reported a recovery of 95 ~ at 4 x 10-7 ~ oleate and by

ISKRA(13) who reported a recovery of 100 ~ at 1 x 10-6 ~ oleate.

These results Indicate that chemisorption takes place at the fluorite surface priorto surface precipitation and that a hydrophobic surface Is created under these

condit 10M. The characteristic features of chemisorption and surface precipitation are

quite different(ll).

Little of the chemisorbed oleate is removed by ultrasonic treatment. However,

the surface precipitate is held rather loosely at the fluorite surface and can be

partially removed by ultrasonic treatment.

Further as. shown in fic. 2. the chemisorption reaction is round to be endothermic

with an increase in adsorption density at higher temperature. On the other hand the

adsorption density in the surrace precipitation rerion decreases with an increase in

temperature as expected rrom the temperature dependence or the solubility or

cal~ium oleate(ll).

132

.,{,f'

In fact the time sequence of photOCl'8Phs for bubble detachment In fiC. Jdemonstrate rather vividly the weak forces which hold the surface precipitate at the

fluorite interface(14).

Such bubble detachment by the bouyant roree does not oceur when the oleate II

che m iIOrbed.

In view of these data. It seems rather clear that a unique collector chemisorptionreaction can occur at mlneraJ surfaces which can be distlncuished from surface

precipitation phenomenon.

The unique character of the chemisorbed layer is due to the fact that partial bond

satisfaction for the collector-salt, chemisorbed species is provided by the surface

states of the crystal lattice.

1. P~k. A.s.. '"Infrared studies of oleic acid ana sodium oleate adsorption onfluorite, barite, and calcite." U.S. Bur. of Mines. Rep. RI 6202, 1963.

2. P~k. A.S.. '"Infrared study of the depression eff~t of fluorite, sulfate, andchloride on chemisorption of oleate on fluorite and barite." Int. Miner. Process.Confre., 7th. New York. Gordon and Breach. Vol. 1,259-67. 1964.

Lovell. V.M.. Goold. L.A. and Finkelstein. H.P.. "lnrrared studiees or theadsorption of oleate species on calcium fluorite," Int. J. Miner. Process. VoL I,113-92,1974.

4. Iskra. J. Kiekowska, M., "Application of ATR technique to fluorite-oleate-quebracho systems," Trans. IMM, Vol. 89-90, C81-92, 1980.

5. Poling, G.W., "Reactions between thlol reagents and sulfide minerals," Inflotation. :\I.C. Fuerstenau. Ed.. Gaudin Mem. Vol. I, pub. AIME. 334-51. 1916.

6. Peck, A.S., Raby, L.H. and Wadsworth, :\f. E., "An infrared study of the flotationof hematite with oleic acid and sodium oleate," Trans. AIM!, Vol. 235, 301-07,1988.

7. Rinelll, G., ~arabini, A.M. and Alesse, V., "Flotation or C!usiterite withsaliC!ylaldehyde as a C!olleC!tor," in flotation, ~.C. Fuerstenau, Ed., Gaudin :\{em.Vol. I, pub. AIME, 549-60, 1976.

8. Pradip and Fuerstenau, D.W., "The adsorption hydroxamate on semi-soluble salts.Part 1 : adsorption on barite, calcite and bastnaesite," ColL Surf., Vol. 8, 108-19,1983.

9. !.tIller. J.D. and !.tisr.. M.. "The hydrophobic character of semi-soluble saltminerals with oleate u coUeetor." mintek 50, international conference on recentadvances in mineral science and technology, Johannesburg. South Africa. "arch1984.

10. Giesekke. P.W. and Harris. P.J.. "A study of the selective flotation of fluoritefrom calcite by the use of a single bubble-stream micro flotation cel"- mintek 50.international conference on recent advances In mineral science and technology.Johannesburg. :'.carch 1984.

11. Hu. J.S., :\tisra, M. and Miller, J.D., "Effect of temperature and oxyren on oleateadsorption by fluorite," to be published in Int. J. Miner. Process., 1985.

l2. Sorensen E.. "On the adsorption of some anionic collectors on fluorite minerals."J. Cou. Inter. Sci. Vol. "5, No. 3, 60l-G7, 1973.

13. Iskra. J.. Gutierrez. C. and Kitchener, J.A.. "FlotatIon of fluorIte. calcite.hematite and quartz with oieate as collector." Trans. (MM. VoL 82. C73-78. 1973.

14. )flller, J.D., Wadsworth, :\foE., )fisra, )f. and Hu. J.S.. "Flotation chemistry or therluorite/ofeate system." in principle or mineral rlotation. the wart symposium.edited by Jones. ~.H. and Woodcock. J. T.. The Australian Institute or Mining' and;Metallurc, 31-42. 1984.

Figure 1 - Effect of ultrasonic treatment on the adsorption isotherms ofoleate on Vent ran optical grade fluorite.T = 20 C. Conditioning time = 2 hours.

134

!

Figure 2 - Adsorption isotherms of oleate on Ventron optical gradefluorite at 20 C and -10 C when using ultrasonic treatment.Conditioning time = 2 hours.

8U88lE DETACHMENT TIME SEQUENCE,-.". 0' 10'. rr" C

Figure 3 - Time sequence photographs for the detachment or an airbubble.

i)s

A. from Pr. P. SOMASUHDARAN

M[LLER et aI. claim, based on past infrared results, that oleate does tnlIJ

chemisorb on semi-soluble minerals.

As we have pointed out several times in the past, previous fa results cannot be

used to draw any conclusions regarding adsorption mechanisms since fa is not an in

situ test and pretreatment such as consoiidation of particles (altering the double

layer interactions) and even more seriously, drying will produce artifices and can

yield misleading results.

During drying, the aqueous film around the particle will ~ome supersaturated

with respect to the surfactant and inorganics, leading to their precipitation.

Spectra exhibited by such precipitates will easily suggest chemisorption if the

presence of the precipitate is not taken into account.

Of course, if the samples are washed prior to the IR test. even more serious

artifices can result.

Three references have been cited for previous work on surface precipitation.

MILLER's reference 8 discusses surface reaction which is different from surface

precipitation as bulk reactions are from bulk precipitation; thermodynamic

requirements are different for precipitation.

References 9 and 10 are unpublished work and we cannot comment on them, norcan we consider as prior work.

In this paper we stress that precipitation conditions can occur in the interfacial

region even before such conditions are met in the bulk solution. Electrostatic andother adsorption forces will playa critical role in creating such a situation.

The pH dependence of metallic cation adsorption on silica is a typical example of

the above type where surface precipitation can be predicted to occur even in the

absence of bulk precipitation.

Results tOI" the adsol"ption ot oleate on tluol"ite given in fig. 1 indeed suggest

pl"ecipitation undel" these conditions.

136

It is not clear from the information provided whether or not any bulk precipitation

occurred u.nder those conditions. An important possibility that needs to be examined

here is whether or not the adsorption at oleate levels below 1 mg/l occurred under

surface precipitation conditions. If flotation recovery was about 90-100 % at about

0.3 mg/l (1 x 10-6 kmoVm3), one would expect the adsorption to have exhibited a

sharp increase with notation at some concentration below 0.3 mg/i.

,

The data in fiCo 1 does not clearly show the adsorption behavior in that low

concentration region.

We feel that the adsorption behavior in that region needs to be examined taking

into account the electrochemical equilibria under those conditions. Note that the

mineral-solution equilibria under those conditions can be markedly different

depending upon whether or not the system was open to atmospheric CO2-

Our calculations show that for a closed system. Ca-oleate can precipitate at

oleate levels as low u 10-6 kmoVm3.

Thus, it is clear that under conditions when MILLER et al. obtained flotation.

precipitation may have already begun to occur on the mineral surface.

The effect of temperature on the behavior of semi-soluble minerals is complex. In

addition to the solubility of calcium oleate and the amount of calcium and other

soluble species released into the system, the electrokinetic properties of the solid-

liquid interface will be affected by temperature.

Therefore, while the thermodynamic solubility product itself may be higher at

higher temperatures, the actual amount precipitated in the surface region will depend

upon many factors that are not understood at present. Thus, it is premature to

speculate at this stage as to how surface precipitation will change with temperature.

Q. rrom Dr. 8.M. MOUOOIL (University or Florida - USA)

Have you done any surface characterization to prove the presence of surface

precipitates, when no bulk precipitation will be expected?

A. from Pr. P. SOMASUNDABAN

We(23,24,S,9} have characterized the surface using electrokinetic measurements.

137

~'A

We believe that the electrokinetic technique Is a powerful toot to characterize the

surface In situ without subjectlnr the sample to the severe pretreatments. such as

drylnr. that are required tor spectroscopic analysis.

The examples discussed in this paper as well as those in reference 19 show that

the electrokinetic behavior ol the surface becomes progressively similar to that ol

the bulk precipitate as the particles are contacted with the surfaetant.

Q. from P.E. RICHARDSON (BUJ'eau of Mines)

I would greatly appreciate it ir Pro SOMASUNDARAN would comment on the

flotation or a system which exhibits both chemisorption and precipitation.

I have in mind chaJcoeite. which as shown by POMIANOWSKI a number of years

ago can chemisorb xanthate. Pro. FUERSTENAU and Dr. CHANDER also have shown

dlthiopftosp!1ate chemisorbs on this mineral. In our laboratory. we have recently

confirmed the" xanthate chemisorption mechanism and shown also that a copper-

xanthate can be precipitated.

At open'-C!ircuit Cu(1() ions are released to the solution phase. Xanthates reaC!t

with Cu(I[) to form a preC!ipitate but this reaC!tlon is detrimental, not benefiC!ial to

flotation.

When does the precipitate mechanism lead to flotation?

A. from Pr. P. SOMASUNDARAN

Invariably, in chemisorption systems, precipitation also can be expected. In fact.

the examples discussed in the paper clearly suRest that true chemisorption itself

may be a surface precipitation phenomenon. In these systems, surfactant will interact

also with the dissolved species in solution and form bulk precipitate. The extent of

bulk vs. surface precipitation will depend essentially upon the mineral solubility and

the dissolution kinetics. Hiih solubility and fast kinetics will promote bulkprecipitation. Results for the copper-oxlme system(ll, 13) as well as for calcite-

apatite-oleate system(1) clearly show that the bulk precipitate does not act as a

collector in this case(13).

The collector consumed (or bulk precipitation Is thus unavailable for flotation andhence bulk precipitation is detrimental for flotation. Precipitation on the mineral

138

I

surface, on Ute other hand, can aid notation provided the orientation of the adsorbate

is appropriate. The Cu-xanUtate system can be expeeted to be somewhat similar to

the Cu-oxlme system and hence both surface and bulk precipitation can occur and

bulk precipitation will then be unfavorable for flotation.

Q. trom h. s. OAIOARJIEV (Mlninc and OeolocicaJ Institute. Sotla - BuJprie)

1. In the presentation, you mentionned C.M.C.

How can this concentration be optimized ?

2. In the presentation and in the answers to previous questions. it has been said that

results are mainly interpreted throurh thermodynamies considerations.

What is the opinion of the author about an interpretation more buic and universal

based on the energy principle?

A. from Pro P. SOMASUNDARAN

( am ~ot quite sure why Pro GAIDARJIEV wishes to optimize C:\tC. However. if

required. it can be optimized by usinr an appropriate combination of hydrophilic

(polar and ionic) groups and hydrophobic (alkyl and aryl) groups in the collector

molecule.

Again, it is not clear as to what differences between the thermodynamic and

energetic considerations are beinl' intended.

(f it is the kinetic consideration that is being intended. indeed. it can have a major

influence particularly if the conditioning time is less than that required for

equilibrium adsorption.

Q. from J.-M. CASES

( am not sure that the words surface precipitation do not correspond to a false

problem.

Ir you had to write the theory or this new concept, will you take into

consideration the laws of adsorption or those of three dimensional condensation on a

surface?

In that eue. eQUId you omit all the well known published examples where the

number of layers is limited to 1 or 2 ?

The theory ot three dimensional condensation on a surface Implies the presence ot

a great number ot layers on the surface. (This number belnl then the necessarly

even).

A. from PI'. P. SOMASUNDARAN

We have shown that in a number of so ealled ebemimrptlon systems. adsorption

may have oecurred under eondltlons for ehemieal preeipitation and. most

Importantly. sueh preeipitation eould oeeur In the interfaeial recton at solution

eoneentrations lower than that for bulk preeipltation.

For example, while bulk ~hemi~al equilibria does not predl~t preeipltation under

the pH ~onditlons (-10) of the sharp in~rease In Ca adsorption on slll~a, ~al~ulations

usinr Ca con~entration In the interfa~ial recton and bulk thermodynaml~ solubility

produ~t show the possibility for pre~ipitation in the interfa~ial region.

True trl-dlmensional condensation (essentially Infinite In three dimensions) often

mlitlt not occur on the surface (thus limiting the precipitation to I or 2 layers) since

the forces such as the electrochemical potential responsible for the surface excess of

reagent required for precipitation can disappear. or at least chanre. as the particle

surface Is coated and thus masked by the precipitate.

Thus. once the silica particles are coated by a few layers of Ca(OH)2 which is not

neptlvely chqed u is the silica at pH 10. electrostatic adsorption of calcium in the

Interfacial rei'ton cannot be expected.

In the absen~e of ex~ess ~aJ~ium in the interfa~ial recion, further pre~ipitation of

Ca(OH)2 to form additional iayers is not expe~ted.

0

0

140