Colloids and Surfaces, 13 (1985) 73-85 Elsevier Science Publishers B. V ., Amsterdam

Printed in The Netherlands

73

A STUDY OF POLYMER/SURF ACTANT INTERACTION AT THE MINERAL/SOLUTION INTERFACE P. SOMASUNDARAN and J. CLEVERDON School of Engineerinl and Applied Science. Columbia Uniuersity, New York. NY 10027 (U.S.A.) (Received 4 October 1983; accepted in final form 18 April 1984)

ABSTRACT Interaction between a cationic polymer (acrylamidemethacrylamidopropyltrimethylammonium chloride copolymer) and cationic and anionic surfactants (dodecylammonium chloride and dodecylsulfonate) on quartz is studied using adsorption, electrokinetic and flotation techniques. At pH 6.5, the polymer depressesamine flotation of quartz but without depressing amine adsorption on it. E]ectrokinetic experiments yielded zetapotentia] values characteristic of the polymer as long as polymer ia adsorbed irreapective of both amine adsorption and original potentia] of quartz. Around pH 10 where amine is most surface active, polymer adsorption from a solution containing both the polymer and amine is negligible. Furthermore, at this pH any preadsorbed polymer is displaced from the surface upon subsequent addition of amine. A molecular model is proposed for the polymer-surfactant layer on the quartz partic]e with the massive polymer species masking the adsorbed amine, to account for the hydrophilic characteristics of the particle. In accord with this model, quartz is activated by the adsorbed cationic polymer for flotation by the anionic sulfonate.

INTRODUCTION

In many industrial processes,such as flotation and enhancedoil recovery, polymers are now used along with surfactants. Flocculation followed by flotation of iron ore is a recent example of such a process [1, 2]. In addition, nowadays polymers very often are added either to aid filtration, to clean effluents or as grinding aids. These reagentscan markedly affect other downstream processessuch as flotation [3-10]. In such operations, selectivity can be affected through interactions between polymer and surfactant in the bulk or at the solid/liquid or liquid/gas interface. At the solid/liquid interface, not only the amount of eachreagent adsorbedbut also the manner (orientation) of adsorption has an important effect on the process.However, to the authors' knowledge, no precise information on these factors exists in the literature, although there have been a few reports on interactions in bulk solutions [11-24]. Recent work [11] on such solution properties as relative viscosity, conductivity, and surface tension with various nonionic, anionic and cationic polyacrylamides and sodium dodecylsulfonate or dodecylamine

0166-6622/85/S03.30

C>1985 Elsevier Science Publiahers B. V.

74

hydrochloride showed measurableinteraction effects but only in the caseof the oppositely charged polymer and surfactant systems. Also depending on the ionic natures of the polymer and surfactant they can precipitate, and in some casesthe precipitate thus formed can dissolveupon increasingthe surfactant concentration. This redissolution was considered to occur through complexation. It was noted that the observed effects can affect the adsorption behavior of different speciesat the solid/liquid interface and thereby influence processes such as flocculation, flotation and micellar flooding. In this work, interactions of a cationic polymer, and anionic and cationic surfactants at the quartz/aqueous solution interface are examined using adsorption and electrokinetic techniques together with flotation. Clearly seen is the power of the use of the electrokinetic technique in conjunction with other techniques in deriving information on the possiblearrangementof adsorbed speciesat the interface. EXPERIMENTAL

Materials High purity crystalline quartz (Arkansas) purchased from Ward's Natural SciencesEstablishment was crushed and ground in a stainless-steelball mill; the -20-lJ.m fraction was cleaned by boiling in HCI solution and repeated rinsing with water, freeze-dried and stored till use. The averageparticle size as measured using a Fisher subsievesizer was 6 IJ.m.For flotation experiments, -590 to +210-lJ.mfraction of quartz was used after cleaning with a magnetic separatorand leachingwith warm nitric acid. The cationic copolymer PAMA used in this study was prepared from 14Clabeled acrylamide and methacrylamidopropylltrimethylammonium chloride by irradiating them in acetone-water solution using 6°Co source at 6 krads h -I for 18 h. Unreacted monomer was removed by repeated washing with acetone and the sample was freeze-dried and stored till use. The molecular weight was estimated from intrinsic-viscosity data to be 0.8 X 106. Dodecylammonium chloride and sodium dodecylsulfonate purchased from Eastman Kodak Co. and Aldrich Chemical Co., respectively, were used as collectors without further purification. Procedure For adsorption tests, 2 g of the - 20-JJmquartz were conditioned first for two hours on a wrist-action shaker in 10 ml of NaC! solution adjusted to the desiredpH, and then for another six hours after adding 10 ml of the polymer solution. (Equilibrium was attained within this interval at all pH levels, Fig. 1). The pH was then measured,the sample was centrifuged and the residual polymer concentration was determined using a liquid scintillation counter. Adsorption of amine was determined by titration againsta known concentra-

n tion of sodium dodecylsulfonate using a two-phase technique [25]. To determine the extent of polymer desorption, after centrifugation 10 ml of the supernatant were replaced with 10 ml of NaCl solution, and the sample agitated for another six hours. pH

0 1.2

0

" 0 E

.

z Q le.

c I

1.0

T

..

.

10.4 j; 0.4

0 3 I 10-2 kmo' 1m3 HoC' 24 t 1- C

SIL. . 10 %

~ 0.8 '" ~ 0.61

CO~C. POL.YMER . 260 IIIQ/kQ 6.0 t; 0.& (natural)

0.4

t; 0.1

£).

0.2 0.8 ..

..

10

12 TIM!.

14 hour..

It.

Fig. 1. Adsorption of PAMA on quartz as a function of time.

For zeta-potential measurements,0.02 g of quartz was conditioned in 40 ml of total solution in the same manner as in the adsorption experiments. Polymer-adsorption data in these caseswas simultaneously determined by centrifuging the samples after zeta-potential measurementsand analyzing the supernatant for residual polymer concentrations. For flotation, 1 g of the -590 to +210ilm quartz was conditioned with the necessaryreagentson a tumbler at 16 rpm for 10 min and then floated in a modified Hallimond tube for 15 s using nitrogen at a glas-flow rate of 36cm3min-1 [3]. RESULTS AND DISCUSSION

PAMAadsorption The extent of adsorption on PAMA on quartz was first determined under natural pH conditions (pH 6.3 j; 0.6). Figure 2 showsmeasurableadsorption at concentrations as low as 1 mg per kg. Adsorption appearsto reach a plateau at about 500 mg per kg. The results of dilution tests show that adsorption of PAMA is essentially irreversible by dilution under the conditions used here. Furthermore, it is seen that, while equilibrium was possibly attained within the six hours of conditioning below 100 mg per kg, above this concentration adsorption continues to take place even after dilution.

76

~ ... ~ E z 2 ... Go ~ 0 III 0 C

01

1.0

10

RESIDUAL

100

CONCENTRATION,

"'v/kv

Fig. 2. Adsorption isotherm of PAM A on quartz under natural pH conditions.

Quartz flotation in amine and PAMA solutions The effect of polymer addition on quartz flotation using amine was tested at two different amine levels. Results given in Fig. 3 show that complete depression of flotation is achieved at 1 mg per kg polymer. In order to determine whether or not this depression of hydrophobicity is due to reduced adsorption of amine (the traditional explanation), adsorption of amine on quartz was determined in the presence and absence of the polymer. The results are tabulated inside Fig. 3. Most interestingly, the presence of the polymer resulted in decreased flotation without any significant effect on the surfactant adsorption. We have made similar observations for the calciteoleate system using starch as depressant [26]. 100 QUARTZ/AMINE/PAMA

80

Q '" ... 4 0 -' ... .t

0

AMINE (kmol/m") 36.10'0

t.

84.10-0 pH. 65

60

40

\.\

20

~

\,

Effect of POlyme' on AmIne Adscrptlon (C, E 63x'O"kmO:/",3)

,

Polymer, . mQ/kQ

~

500

46144

0: 0

I

10

100

POLY~ERCONCENTRATION. m9/k9 Fie. 3. Depreuion of notation of quartz using dodecylamine by the cationic polymer PAMA at natural pH. Adsorption of amine in the pre-nce and absenceof the polymer is shown in the inset.

77

Zeta potential in amine solutions To obtain additional information about the nature of the mineral surface under these conditions, the zeta potential of the -20-p.m quartz particles was determined together with polymer adsorption, at different polymer and amine concentrations, as a function of pH. Ionic strength was maintained at 3 X 10-2 kmol m-3 using NaCI for all tests. Figure 4 showsthe pH dependence of quartz zeta potential in 3 X 10-2 klnol m-3 NaCI and 5 X 10-4 kmol m-3 amine plus 3 X 10-2 kmol m-3 NaCI solutions. An isoelectric point of about pH 3.0 is obtained for quartz, in agreement with literature values which vary from 1 to 3.7 [27]. The presenceof amine is found to have a significant effect on zeta potential with charge reversal in a narrow pH region around 10. Notably, this is also the pH region in which the highest quartz flotation is usually obtained [28]. This phenomenon has been explained by taking into account the formation of the highly surface-activeamineaminium complex in this pH range (see Fig. 5). At higher pH values most of the amine is in the neutral form which does not lend itself to adsorption [29] .

Fig. 4. Zeta potential of quartz in water and in dodecylamine solution as a function of pH, both under constant ionic-strength conditions.

78 ~DDDECYLAMINEHYDROCHLORIDE C . ~'10-4kmOl/m3

2

,0

8 pH

RNH;

. 10.63 .

RNH;

~

RNH2 + H+

pKa

2RNH;

~

(RNH3)~+

pKO

+ RNH2

~

(RNH2RNH3)+

pKAO'

- 3.12

RNH2(.(J ~

RNH2(aq,)

PKS'

4.69

-2.08

I'. + 2C CT . CR"M, + CRuM~ 3 + 2t"-RNM , 32 (~"M 2RNM3I. Fig. 5. Dodecylamine species of 5 X lO--kmol mos.

diagram

as a function

of pH for a total

amine

concentration

Zeta potential in polymer solutions and polymer adsorption Zeta potential of quartz particles at 0.1, 0.4, 0.8, and 2.5 mg per kg polymer addition is given in Figs. 6-9 as a function of pH together with adsorption at the three higher dosages.Even at only 0.1 mg per kg polymer addition, the zeta potential is altered markedly at all pH values above 2, and at

It

~ ".1

.. z ... .. 0 Q.

~

~

--.J

..,'

---"'

'.

~

J

I

I

J

J.-

CJ

C -10 .. ... N

I:

T

-20

()

.

3"O-2""'OI/,..3NOCI

2~ i '.C

S/L . 005%

CONCPOLYMER. 0' ~,

(J

"'9/"9 .,

56789101112

,

.

,

,

,

I

pH

Fig. 6. Zeta potential of quartz as a function of pH at 0.1 mg per kg PAMA dosageunder constant ionic-strength conditions.

~

79

0.4 mg per kg, it reachesa value of about +12 mV which shows very little variation with pH and further addition of polymer. On the other hand, polymer adsorption increaseswith both pH and polymer dosage.The increaseof adsorption with pH suggeststhat electrostatic forces playa significant but not exclusive role in the adsorption of this polymer on the quartz surface. Electrostatic force is not considered to be the exclusive controlling parameter because,even at the isoelectric point of quartz, there is a measurable effect of polymer adsorption on zeta potential even though the adsorption

.,

. I - 3010-2 kmol/m3NoCI r-24J.l.C S/L - 005"1. CONC POLYMER.

0.4 mg/kg

~ E

-- -- -'-'

~

too Z W too

'-

~~~~~~~~~~~

~ E

~~~~~,~~

f c .10

too w N

..

-.-

-30'

7

-

--.M-

-,*",

-

-

r:~'..,

.

.

,

.

I

,I

3

4

5

6

7

8

9

I

'-

10

11

>-t: VI Z &oJ 0 Z 0 1 -: Go ~ 0 VI 0 d

12

pH

Fig. 7. Zeta potential of quartz and polymer adsorption on it as a function of pH at 0.4 ml per kl PAMA do..,e under constantionic-strenlth conditions. +30 I.3'10-2k",OI/m3NoCI T.25.tI.C .,01

.

S/L

O.b5%

CONC POLYMER.

..,...().G-.=L-.

~ .10

c("

-' C

~

'" ... a Q.

._.:-::--~

;:...

0.8 "'9/kg

e oJ ~ ... z w ... i

~ W N

Fil. 10. Zeta potential of quartz and polymer adsorption u a function of pH in solutions containinl both the polymer (2.5 D1I per kg) and dodecylamine (5 x 10-4 kmol mol). TABLEt Compariaon of zeta potential and ad80rption in quartz/amine, quartz/PAMA, and quartz/ amine/P AMA systems pH

Zeta

Quartz/amine

6 9.6 10.2 10.4

-6.5 +16.5 +26.8 +26

Quartz/PAMA

6 9.6 10.2 10.4

+10 +9.5 +10 +10

1. 2..4

Quartz/amine/p AMA (amine and PAMA added limultaneoully)

6 9.6 10.2 10.4

+9 +25 +27 +23

1..4 0 0 0

Quartz/amine/PAMA (polymer added rlrSt)

6 9.6 10.2 10.4

+17.5

0

+21.5

0

rpAMA

-

.1

rAmiDe

0.052 1.05 1.05 1.05

-

2 1,.95

0.072

0.106 1.17 1.17 1.17

82

Noting that the polymer PAMA did adsorb on quartz in alkaline solutions but did not when amine was present, tests were conducted to determine whether the amine can possibly causedesorption of the polymer around pH 10 where amine is most surface active. Theseexperiments entailed conditioning of the mineral first with the polymer and then with amine. If polymer adsorption is irreversible, when amine is added after the polymer hu already adsorbed, the zeta potential should be about +10 mV, the value obtained in the absenceof amine. However, the zeta potential results at pH 9.6 and 10.4 are +17.5 and +21.5 mV, respectively (Table 1). Also, the polymer adsorption in both tests is zero. Thesezeta-potential valuesare similar to the values obtained with amine alone (some difference results from the extreme sensitivity of the zeta potential of quartz in amine solutions around pH 10). Since -

Fie. 11. (a) Schematic reprelentation or the cationic polymer PAMA and dodecylamine co-adlOrption on quartz particles resultina in their notation depreuion. (b) Schematic reprelentation or quartz/dodecylaulronate system. (c) Schematic representation or the cationic polymer PAMA and dodecylaulronate co-adsorption on quartz particles resultine in their notation activation.

84

Electrokinetic experiments conducted together with adsorption proved to be a powerful method of studying the nature of polymer-surfactant interactions that are responsiblefor the observedeffects on particle wettability of these reagents. While the presence of amine alone produced increasing changes in the zeta potential with increasing pH, with maximum effect around pH 10, polymer addition alone even at a dosageas low as 0.4 mg per kg yielded a constant value of about +10 mV (possibly characteristic of the quartz surface masked by polymer) in the complete pH range. Even when the polymer and the surfactant were present together in the system, a zeta potential near +10 mV was obtained in the acidic and neutral pH regions where polymer adsorption had occurred. Noting that the amine adsorption under these conditions did not make the quartz particles hydrophobic and that it did not influence the zeta potential, a model is proposed where the massive polymer speciesadsorb on the amine-eoatedparticles masking the adsorbed amine. This model is supported by the activation by PAMA observed for sulfonate adsorption on quartz particles and their flotation. ACKNOWLEDGEMENT The authors acknowledge the Chemical and ProcessEngineering Division of the National ScienceFoundation for support of this work. REFERENCES 1 J.W. Villar and G.A. Dawe, Min. Coner. J., 61 (1975) 40. 2 P. Somasundaran, in P. Somasundaran (Ed.), Beneficiation of Mineral Fines, AIME, New York, NY, 1979, pp. 183-196. 3 P. Somasundaran and L.T. Lee, Sep. Sci. Technol., 16 (1981) 1475. 4 P. Somasundaran and L.T. Lee, Polymer-Surfactant Interactions in the Flotation of Quartz and Hematitie, International Mineral PrOcelSingCongress, Preprints, 1982, IV 9.1-IV 9.17. 5 BM. Moudgil and P. Somasundaran, Adsorption of Charged and Uncharged Polyacrylamides on Hematite, Society of Mining Engineers, 1982, Preprint, pp. 82-160. 6 P. Somasundaran, in P. Somasundaran (Ed.) Fine Particles ProceasiDi, Vol. 2, AIME, New York, NY, 1980, pp. 946-976. 7 L. Usoni, G. Rinelli, A.M. Marabini and G. Gbiii, Selective Properties of Flocculants and Possibilities of Their Use in Flotation of Fine Minerals, VIIIth International Mineral ProcessingConrress, Leningrad, 1968, paper D-13. 8 G. Ghili and C. Botre, Trans. Inst. Min. Metall. Sect. C, 75 (1966) 240. 9 G. Gbigi, Trans. Inst. Min. Metall. Sect. C, 78 (1968) 212. 10 J. Szczypa and S. Chibowaki, Colloids Surfaces, 3 (1981) 393. 11 P. Somasundaran and B.M. Moudgil, in R.B. Seymour and G.A. Stahl (Eds.), Macromolecular Solutions, Pergamon, New York, 1982, pp. 151-165. 12 J. Rubio and J.A. Kitchener, Trans. Inst. Min. Metall. Sect. C, 86 (1977) C97. 13 S. Saito and Y. Mizuta, J. Colloid Interface Sci., 23 (1967) 604. 14 M.J. Schwuger, J. Colloid Interface Sci., 43 (1973) 491. 15 E.D. Goddard and R.B. Hannan, in K.L. Mittal (Ed.), Micellization, Solubilization and Microemulsions, Plenum, New York, NY, 1977, pp. 835-845.

88

these "polymer first - amine second" tests yielded a value for zeta potential close to that obtained when both reagentswere added simultaneously. it is possible that in this casethe polymer molecule was displaced by the competing amine-aminium complex. This suggeststhat polymer desorption can be induced by appropriate competing species.Similar desorption of hydrolyzed polyacrylamide from clay has been achievedrecently using phosphates[30]. Whereasdesorption was not induced by mere dilution (Fig. 1), in this case desorption is possible since sites vacated by individual segmentscan now be occupied by the competing speciesresulting eventually in the desorption of the entire polymer molecule. On the basis of the model proposed here, it should be possible to float quartz using an anionic surfactant even though nonnally quartz will not respond to an anionic surfactant (Fig. 11b,c). Flotation experiments conducted with dodecylsulfonate (together with PAMA) showed that quartz can indeed be floated completely with this reagent combination (Fig. 12) [4]. Adsorption experiments showed quartz to extract the sulfonate from the solution but only in the presenceof PAMA (seetable in Fig. 12). In addition to interactions of the polymer and the surfactant on the particle surface. interactions in the bulk liquid [11] also can contribute to the activation of quartz by PAMA for flotation using the anionic surfactant. but they are not consideredto be primarily responsiblefor the observedflotation.

0

, POLYMER

10

100

CONCENTRATION.mQ/kQ

Fig. 12. Activation of notation of quartz u8ing dodecylaulfonate by the cationic polym@r PAMA at natural pH. AdlOrption of 1U1fonatein the pre8@nceand abeenceof the polymer ia shown in the inlet.

Concluding remarks Flotation of quartz by amine is depressedby the cationic polymer PAMA, while the amine adsorption itself is not affected.

84

Electrokinetic experiments conducted together with adsorption proved to be a powerful method of studying the nature of polyme~urfactant interactions that are responsiblefor the observedeffects on particle wettability of these reagents. While the presence of amine alone produced increasing changes in the zeta potential with increasing pH, with maximum effect around pH 10, polymer addition alone even at a dosageas low as 0.4 mg per kg yielded a constant value of about +10 mV (possibly characteristic of the quartz surface masked by polymer) in the complete pH range. Even when the polymer and the surfactant were present together in the system, a zeta potential near +10 mV was obtained in the acidic and neutral pH regions where polymer adsorption had occurred. Noting that the amine adsorption under these conditions did not make the quartz particles hydrophobic and that it did not influence the zeta potential, a model is proposed where the massive polymer species adsorb on the amine-coated particles masking the adsorbed amine. This model is supported by the activation by PAMA observed for sulfonate adsorption on quartz particles and their flotation. ACKNOWLEDGEMENT The authors acknowledge the Chemical and ProcessEngineering Division of the National ScienceFoundation for support of this work. REFERENC~ 1 J.W. Villar and G.A. Dawe, Min. Contr. J.,61 (1975) 40. 2 P. Somuundaran, in P. Somuundaran (Ed.), Beneficiation of Mineral Fines, AIME, New York, NY,1979, pp.183-196. 3 P. Somuundaran and L.T. Lee, Sep. Sci. Technol., 16 (1981) 1475. 4 P. Somaaundaran and L. T. Lee, Polymer--Surfactant Interactionl in the Flotation of Quartz and Hematitie, International Mineral Proceuine Conrreu, Preprintl, 1982, IV 9.1-IV 9.17. 5 B.M. Moudeil and P. Somuundaran, Adsorption of Charged and Unchareed Polyacrylamidea on Hematite, Society of Minine En,meen, 1982, Preprint, pp. 82-160. 6 P. Somuundaran, in P. Somaaundaran (Ed.) Fine Particles Proceuine, Vol. 2, AIME, New York, NY, 1980, pp. 946-976. 7 L. Usoni, O. Rinelli, A.M. Marabini and G. Ghiei, Selective Propertiea of FJocculantl and Pouibiliti.. of Their Use in Flotation of Fine Minerall, VllIth International Mineral Proceuine Contreu, Leninerad, 1968, .-per D-13. 8 G. Ghili and C. Botre, Trans.lnlt. Min. Metall. Sect. C, 75 (1966) 240. 9 G. Ghili, Trans. Inat. Min. MetalJ. Sect. C, 78 (1968) 212. 10 J. Szczypa and S. Chibowlki, Conoids Surface., 3 (1981) 393. 11 P. Somuundaran and B.M. Moud(i), in R.B. Seymour and G.A. Stabl (Eds.), Macromolecular Solutionl, Pereamon, New York, 1982, pp. 151-165. 12 J. Rubio and J.A. Kitchener, Trani. Inlt. Min. Metall. Sect. C, 86 (1977) C97. 13 S. Saito and Y. Mizuta,J. Colloid Interface Sci., 23 (1967) 604. 14 M.J. Schwueer, J. Colloid Interface Sci., 43 (1973) 491. 15 E.D. Goddard and R.B. Hannan, in K.L. Mittal (Ed.), Micellization, Solubilization and MicroemuJaiona,Plenum, New York, NY, 1977, pp. 835-845.

85 16 17 18 19

20 21 22 23 24 25 26 27 28 29 30

M.N. Jones, J. Colloid Interface Sci., 23 (1967) 36. H. Arai, M. Mukata and K. Shinoda, J. Colloid Interface Sci., 37 (1971) 223. E.D. Goddard and R.B. Hannan, J. Am. Oil Chem. Soc., 54 (1977) 561. E.D. Goddard and R.B. Hannan, J. Colloid Interface Sci., 55 (1976) 73. K.E. Lewis and C.P.Robinson,J. Colloid Interface Sci., 32 (1970) 539. S. Saito, J. Colloid Interface Sci., 23 (1967) 227. B. Cabane, J. Phys. Chem., 81 (1977) 1639. B. Kalpakci and R. Nagarajan, Surfactant Binding to Polymer and Phase Separation in Aqueous Surfactant-Polymer Solutions, AIChE Meeting, Houston, April, 1981. N.J. Turro, B.H. Baretz and Ping.Liu Kuo, Macromolecules, 17 (1984) 1321. G.W. Powers, The Volumetric Determination of Organic Sulfonatea by Double Indicator Method, Communication C.225, Amoco Production Co., Tulsa, 1970. P. Somaaundaran, J. Colloid Interface Sci., 31 (1969) 557. G.A. Parks, Chem. Rev., 65 (1965) 177. K.P. Ananthapadmanabhan, P. Somuundaran and T.W. Healy, Trans. AIME, 266 (1980) 2003. P. Somuundaran and K.P. Ananthapadmanabhan, in K.L. Mittal (Ed.), Solution Chemistry of Surfactanta, Plenum, New York, NY, vol. 2, 1979, pp. 777-800. P. Dodson and P. Somuundaran, J. Colloid Interface Sci., 97 (1984) 481.