The aqueous catanionic system sodium perfluorooctanoate–dodecyltrimethylammonium bromide at low concentration

Journal of Colloid and Interface Science 312 (2007) 425–431www.elsevier.com/locate/jcis

The aqueous catanionic system sodiumperfluorooctanoate–dodecyltrimethylammonium bromide

at low concentration

José Luis López-Fontán a, Elena Blanco a, Juan M. Ruso a, Gerardo Prieto a, Pablo C. Schulz b,∗,Félix Sarmiento a

a Departamento de Física Aplicada, Facultad de Física, Universidad de Santiago de Compostela, Spainb Departamento de Química, Universidad Nacional del Sur, Bahía Blanca, Argentina

Received 12 January 2007; accepted 18 March 2007

Available online 24 March 2007

Abstract

The interaction between sodium perfluorooctanoate (SPFO) and dodecyltrimethylammonium bromide (DTAB) was studied by several methodsand it was found strongly synergistic. Above a mole fraction of SPFO in the surfactant mixture (αSPFO) = 0.38, the interaction is repulsiveand increases with the content of SPFO in both, the overall mixture and micelles, whereas the interaction is attractive if DTAB is in excess. AtαSPFO = 0.38 the low miscibility between hydrocarbon and fluorocarbon is counterbalanced by the electrostatic attraction between the oppositecharged head groups, and the micelle composition is ideal (i.e., the mole fraction of SPFO in micelles XSPFO = αSPFO = 0.38). The solubility offluorocarbon in hydrocarbon is lower than that of hydrocarbon in fluorocarbon. Micelles of DTAB act as a solvent for SPFO without importantstructural changes, whilst micelles of SPFO undergo important changes when dissolve DTAB. This asymmetry may be interpreted as caused bythe difference in chain length that favors the inclusion of the shorter chain in micelles of the longer surfactant, but disfavors the opposite process.Above XSPFO = 0.5 there is an excess adsorption of bromide ions on the mixed micelles surface, giving rise to a high ζ potential. Micelles of pureSPFO or pure DTAB show an important energy barrier which prevents micelle flocculation. The inclusion of SPFO in DTAB micelles produces areduction of the energy barrier, which disappeared when αSPFO = 0.5. This produces the flocculation of micelles giving rise to the formation of anon-birefringent coacervate, which is probably formed by unordered isometric clusters of micelles.© 2007 Elsevier Inc. All rights reserved.

Keywords: Fluorocarbon/hydrocarbon mixed surfactants; Catanionic surfactants; Coacervate; Sodium perfluorooctanoate; Dodecyltrimethylammonium bromide

1. Introduction

The fluorinated surfactant–hydrocarbon surfactant mixturesare of great theoretical interest for studying the formation andstructure of micelles. Fluorocarbon groups are not only hy-drophobic but oleophobic as well. The phobic interactions be-tween fluorinated and hydrocarbon chains in surfactant mix-tures produce nonideal behavior and, in some systems, demix-ing.

Fluorocarbons have low affinity towards hydrocarbons; thesolubility of fluorocarbons in hydrocarbons is limited, and their

* Corresponding author.E-mail address: [email protected] (P.C. Schulz).

0021-9797/$ – see front matter © 2007 Elsevier Inc. All rights reserved.doi:10.1016/j.jcis.2007.03.035

solutions are highly nonideal. In general a mixture of a fluo-rocarbon and a hydrocarbon surfactant with the same type ofpolar head group shows a positive deviation from ideal behaviordue to the lack of affinity between the two hydrophobic chains.On the other hand, strong coulombic attractions between thehead groups of cationic and anionic hydrocarbon surfactants al-ways lead to very large attractive deviations from ideality ofthese mixed surfactant systems [1]. This is also true for an-ionic fluorocarbon surfactant and cationic hydrocarbon surfac-tant mixtures, even though there is a lack of affinity betweenthe two types of hydrophobic chains. Thus, the study of theeffect of these two opposite forces as a function of the com-position of a mixture of a fluorocarbon anionic surfactant anda hydrocarbon cationic surfactant may give information about

426 J.L. López-Fontán et al. / Journal of Colloid and Interface Science 312 (2007) 425–431

the relative intensities of both effects, namely, the hydrocarbon–fluorocarbon mutual low solubility and the anionic–cationic at-traction.

There are few articles on mixtures of fluorocarbon/hydrocar-bon surfactants having opposite sign of charge [2–4]. Here westudied the aggregation of the catanionic mixture sodium per-fluorooctanoate (SPFO)–dodecyltrimethylammonium bromide(DTAB) to analyze the superposition of the electrostatic at-traction at the Stern micelle layer and the hydrophobic corerepulsion between hydrocarbon and fluorocarbon tails.

2. Experimental

Sodium perfluorooctanoate (SPFO, CAS 1984-06-1) wasobtained from Lancaster Synthesis Ltd. (No. 16988). Dode-cyltrimethylammonium bromide (DTAB) was from Aldrich.Both surfactants were employed as received.

Solutions with mole fraction of SPFO (on a surfactant—onlybasis, i.e., without considering water) αSPFO = 0 (pure DTAB),0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9 and 1 (pure SPFO)were studied. All solutions were prepared by weight. Degassed,double-distilled water was used.

Conductivity measurements were performed with an immer-sion cell and an automatic conductometer, namely Kyoto Elec-tronic C-117 conductivity meter. A series of surfactant solutionsof known concentration was made four days before each mea-surement run. Solutions were vigorously shaken during con-ductivity measurements in a thermostatic bath at 25 ◦C. Thespecific conductivity (κ) data were analyzed as the differencebetween the experimental data (κexp) and those extrapolatedfrom the very diluted solutions (κextrap): �κ = κexp − κextrap.This method enables to see small changes in the conductivitybehavior and increases the slope changes.

Zeta potentials were obtained with a Zetamaster Model 5002(Malvern Instruments, England) by taking the average of (atleast) five measurements at stationary level. The cell used wasa 5 mm × 2 mm rectangular quartz capillary.

3. Theory

3.1. The regular solution theory for mixed micelles

The regular solution theory has been very widely used tomodel the thermodynamic nonidealities of mixed micelles; ithas been shown to accurately model critical micelle concen-tration (CMC) values [5] and monomer-micelle equilibriumcompositions [6] in surfactant systems exhibiting negative de-viations from ideality. However, it must be pointed out that thetheoretical validity of using regular solution theory to describ-ing nonideal mixing in mixed surfactant micelles has been ques-tioned [7]. Although this theory assumes that the excess entropyof mixing is zero, it has been demonstrated that in some sur-factant mixtures this assumption is not true [8,9]. However, thepseudophase separation model and regular solution theory com-bination remains as a very widely used and convenient methodfor analyzing experimental data.

A mixture of two different surfactants 1 and 2 form mi-celles with composition X1 and X2, in equilibrium with solu-tion monomers of composition α1 and α2. These mole fractionsare on a surfactant-only basis, so that

(1)X1 + X2 = 1,

(2)α1 + α2 = 1.

The commonly used equations of the regular solution theoryare presented in Supplementary material. However, some lessknown equations are shown here: accordingly the regular solu-tion theory, the enthalpy of mixing can be computed by [10]

(3)�Hmix = (X1VFC + X2VHC)φFCφHC(δFC − δHC)2,

where VFC and VHC are the molar volumes of the fluorocarbonand hydrocarbon chains, φFC and φHC their volume fraction andδFC and δHC their solubility parameter. Together with the ex-cess free energy of mixing �Gmix the entropy of mixing can becomputed by

(4)�Smix = �Hmix − �Gmix

T.

To compute the enthalpy of mixing the volume fraction ofeach component in the micelle core was computed by

(5)φFC = X1VFC

X1VFC + X2VHC, φHC = 1 − φFC.

3.2. The Derjaguin–Landau–Verwey–Overbeek (DLVO)theory

The theory of the stability of lyophobic colloids indepen-dently developed by Derjaguin and Landau [11] and Verweyand Overbeek [12] is known as the DLVO theory. Since this is awell-known theory, the equations here used are shown in Sup-plementary material.

4. Results and discussion

4.1. The intramicellar interaction

Fig. 1 shows one of the �κ vs C resulting graphs. It maybe seen that three slope changes occurred. The horizontal line

Fig. 1. �κ vs concentration of the SPFO/DTAB mixture with αSPFO = 0.4.The three critical concentrations are indicated. Lines are eye guides.

J.L. López-Fontán et al. / Journal of Colloid and Interface Science 312 (2007) 425–431 427

Fig. 2. The critical concentrations as a function of αSPFO. Lines are eye guides.

corresponds to the monomeric solution. At the concentrationC1 premicelles form, at C2 premicelles transform into micelles,and at C3 a coacervate appears, which was clearly visible as aslightly gray turbidity. This coacervate was slightly denser thanwater and was optically isotropic between crossed polaroids,even when forced to flow. The coacervate did not appear in pureSPFO or DTAB solutions. A characteristic of mixed catanionichydrocarbon–fluorocarbon surfactant mixtures is the formationof large aggregates at high surfactant concentrations. These ag-gregates may actually be tiny droplets dispersed in the aqueousphase instead of being large micelles, which by definition ex-ist in a homogeneous solution [2]. These aggregates have largerdensity than water due to the presence of the fluorocarbon sur-factant. Actually, the formation of coacervates and precipitatesin mixed anionic hydrocarbon–cationic hydrocarbon surfactantsystems is quite common [13–15].

Fig. 2 shows the dependence of the tree critical concentra-tions on the surfactant mixture composition αSPFO. This plotshows that the interaction SPFO–DTAB is not ideal. The idealCMC for the mixture was computed from the regular solutiontheory of mixed micelles with Eq. (13) of Supplementary ma-terial.

The formation of premicelles occurs at concentration higherthan that of pure SPFO at αSPFO � 0.8 that may reflect a lowerconcentration of SPFO.

The CMC data were processed with the regular solution the-ory of mixed micelles to obtain the mole fraction of each com-ponent in the mixed micelles (XSPFO and XDTAB), the mixedmicelle interaction parameter β , the excess free energy of mix-ing and the activity coefficient of both surfactants in micelles,γSPFO and γDTAB.

Fig. 3 shows the composition of micelles (XSPFO) as a func-tion of the mixture composition (αSPFO). It may be seen thatabove αSPFO = 0.38 micelles are systematically richer in DTABthan the ideal composition, and below this composition, arericher in SPFO. The extrapolation from solubility data of n-C7F16 in different normal hydrocarbons to dodecane [10] gavean estimation of the maximum mole fraction of the fluorinatedchain in the hydrocarbon medium of XF ≈ 0.66, which is themaximum mole fraction of SPFO in micelles. The micelle coreis liquid in nature, but its structure is different from that of bulkhydrocarbon/fluorocarbon mixtures, because of the anchoringof one extreme of each chain to the micelle surface. Taking into

Fig. 3. The composition of micelles as a function of αSPFO. Full line is the idealcomposition. Dashed line is an eye guide.

Fig. 4. The interaction parameter in mixed micelles as a function of XSPFO.The dashed line is the last squares fitting straight line.

account this difference, it may be concluded that all micelleswith the exception of those belonging to the system havingαSPFO = 0.9 were not hydrocarbon–fluorocarbon mutually sat-urated solutions.

The mixed micelle interaction parameter β is represented inFig. 1 in Supplementary material as a function of αSPFO, and inFig. 4 as a function of XSPFO. As it occurs in several systems,β was not constant and reflects the changes in the interactionenergy when the micelle composition changes. It may be seenthat above αSPFO = XSPFO = 0.38, the interaction is repulsiveand increases with the content of SPFO in both, the mixture andmicelles, whereas the interaction is attractive if DTAB is in ex-cess (XSPFO � 1 − 0.38 = 0.62). In particular, the dependenceof β on XSPFO was linear (Fig. 4). These results indicate thatthe hydrocarbon chains solubility in fluorocarbon ones is lowerthan that of fluorocarbon chains in hydrocarbon ones. This maybe caused by the fact that the intermolecular attractive inter-action in liquid n-C7F16 is weaker than in n-C7H16. This is ageneral feature for fluorocarbons and hydrocarbons [16]. As aconsequence, at XSPFO = 0.38 the low affinity between fluoro-carbon and hydrocarbon is counterbalanced by the electrostaticattraction and micelles have the same composition that the over-all surfactant mixture.

428 J.L. López-Fontán et al. / Journal of Colloid and Interface Science 312 (2007) 425–431

Fig. 5. The activity coefficients of both components in the mixed micelles as afunction of the micelle composition. Lines are eye guides.

Guo et al. investigated a mixture of sodium perfluorooc-tanoate (SPFO) with n-octyltrimethylammonium bromide [2]and also found a strong attractive interaction.

Fig. 5 shows the dependence of the activity coefficient ofboth surfactants in micelles. The micellar DTAB activity co-efficient γDTAB is almost constant and equal to that of pureDTAB micelles up to XSPFO = 0.38. Actually, there is a smallreduction of γDTAB which indicates a slight attraction. Thismeans that below XSPFO = 0.38, DTAB micelles act as a sol-vent which is not significantly perturbed by the solubilizationof perfluorated chains. Between XSPFO ≈ 0.40 and 0.60 thereis an almost linear increase in γDTAB, followed by a suddenincrease at XSPFO > 0.60. The behavior of γSPFO is different:it slightly increases when DTAB is solubilized in SPFO mi-celles and then remains constant with the increase of DTABcontent up to XSPFO ≈ 0.46. This indicates that SPFO micellesact as solvent for DTAB molecules, but its structure is signif-icantly altered by the presence of the hydrocarbon surfactant.For XSPFO � 0.46, γSPFO decreases linearly, indicating an at-traction between both components. This may indicate that theelectrostatic interaction is stronger than the low affinity betweenthe hydrophobic dislike chains.

To estimate the thermodynamic quantities �Gmix, �Hmixand �Smix we used the values for the normal hydrocarbon (n-C12H26) and fluorocarbon (n-C7H16); VFC = 226 cm3 mol−1,VHC = 214 cm3 mol−1, δFC = 6.0 (cal cm−3)1/2, δHC = 7.879(cal cm−3)1/2 (this value was obtained by extrapolation fromother normal hydrocarbons) [10]. Fig. 6 shows the values ofthe three thermodynamic quantities of mixing. It can be seenthat �Gmix becomes negative at XSPFO < 0.38, and �Smix atXSPFO > 0.46. This indicates that micelles with low SPFO con-tent are thermodynamically favored, whereas those with highSPFO content are not, and that this effect is mainly entropydriven, since enthalpy does not change significantly with themicelle composition.

The interpretation is that between XSPFO = 0 and 0.38,the micelles of DTAB act as solvent dissolving SPFO mole-cules without significantly altering their structure, e.g., thehydrocarbon–hydrocarbon contact, by the introduction of thefluorocarbon chains. Between XSPFO = 0.38 and 0.46 thereis a change in the nature of the micellar mixture, the mutuallow solubility becomes significant and both components be-

Fig. 6. The excess thermodynamic functions of mixing as a function of thecomposition of the mixed micelles. Lines are eye guides.

have nonideally. Above XSPFO = 0.46, SPFO acts as solvent ina mixture with strong non-affinity between components. Then,for XSPFO > 0.46 the micelle core is essentially a fluorocar-bon one with DTAB as solute, whereas for XSPFO < 0.38 themicelle core is of hydrocarbon nature with SPFO moleculesas solute. Between these two compositions, there is the tran-sition between the two extreme situations. The fact that thischange did not occur at XSPFO = 0.5 is caused by the differ-ent strength of the two causes of nonideality: the mutual lowsolubility between hydrocarbon and fluorocarbon in the micellecore, and the Coulomb attraction between the two oppositelycharged head groups.

It is a matter of controversy whether fluorocarbon and hy-drocarbon surfactants mix in their micelles completely or par-tially. Complete non-mixing of a fluorinated surfactant and ahydrocarbon surfactant, partial mixing and the coexistence oftwo types of mixed micelles or one mixed micelle has beenproposed [17]. The above analysis suggests that true fluoro-carbon/hydrocarbon mixed micelles are formed in this system.Guo et al. also found that only one type of micelles is formedin the aqueous SPFO–octyltrimethylammonium bromide sys-tem [2].

The origin of the asymmetry in the solubility of DTAB mole-cules in SPFO micelles and vice versa may be the difference inchain length. In a recent paper [18] the effect of the differencebetween the chain length in mixtures of homologous surfactantswas investigated. A difference in five carbon atoms correspondsto β ≈ −2, and this attraction is attributed to the inclusion ofthe shorter surfactant in the palisade of the longer surfactantmicelle core, thus reducing the hydrophobic core/water inter-face. So, the inclusion of the C7F15-tails in the DTAB micelleswill stabilize the mixed micelles by reduction of the hydrocar-bon/water contact at the micelle surface, whereas the inclusionof the longer C12H25-tails into the small SPFO micelles onlymay be produced by folding the hydrocarbon tail and thus byincreasing the hydrophobic/water interface. As a consequence,SPFO must be more soluble in DTAB micelles than DTAB inSPFO micelles. This is in agreement with the evolution of theexcess entropy of mixing with the micelle composition.

Since the proportion of SPFO molecules dissolved intoDTAB micelles that causes the minimum contact between the

J.L. López-Fontán et al. / Journal of Colloid and Interface Science 312 (2007) 425–431 429

Fig. 7. The total micelle ionization degree vs XSPFO. The curve is an eye guide.

Fig. 8. The true micelle ionization degree vs XSPFO.

core and water is not necessarily when XSPFO = 0.5, this maybe the cause that the change in micelle structure occurs betweenXSPFO = 0.38 and 0.46, and not at the 1:1 proportion.

4.2. The ionization degree

The ionization degree of mixed micelles was computed asα = SM/Sm, where SM and Sm are the slopes of the κ vs C

curve in the micellar and the monomeric zone, respectively.The computed α values are represented in Fig. 7. This α valuetakes into account all the counterions released by micelles. Partof these counterions is released by the formation of ion pairsPFO.DTA, and the other part is formed by counterions releasedby the ionization of the polar head groups of the remaindersurfactant in excess in micelles. The fraction of ionic pairs inmicelles (Nip/n, where Nip is the number of ion pairs in amicelle having the aggregation number n) is equal to the micel-lar mole fraction of the component which is in minor propor-tion, Xmin. Since each ion pair released two counterions whenformed, the true ionization degree of micelles (i.e., that causedby the ionization of the surfactant molecules not neutralizedby an opposite charged surfactant molecule in mixed micelles),αtrue = α −2Xmin. Fig. 8 shows the plot of αtrue as a function ofXSPFO. For the pure surfactants we took α = 0.36 for DTAB at25 ◦C [19] and α = 0.409 for SPFO at 25 ◦C [20]. In this figureαtrue is negative for XSPFO = 0.511 and XSPFO = 0.602. Thismeans that some counterions adsorb to the micelle surface inexcess over the surfactant ions which are not forming ion pairs.

Fig. 9. Zeta potential at the CMC as a function of XSPFO. The curve is an eyeguide.

Fig. 2 in Supplementary material shows the micellar ζ valuesexperimentally determined at the CMC as a function of αSPFO,and Fig. 9 depicts the same potential as a function of XSPFO.It can be seen that the (negative) value of ζ between XSPFO =0.47 and 0.68 is higher than that for pure SPFO micelles. Thismeans that the surface of micelles has an excess of negativecharges, which must arise from the adsorption of bromide ions.This is coherent with the negative values of αtrue. The nature ofthis adsorption is obviously not electrostatic but caused by vander Waals forces because of the high polarizability of Br− ions.

Another feature in the ζ potential behavior is that the Sternlayer in DTAB-rich micelles seems to be very similar to thatof pure DTAB micelles, whilst that of SPFO micelles is verydifferent of both, pure DTAB and pure SPFO micelles. Thisis coherent with the conclusions obtained in the intramicellarinteraction study on the mixed system.

4.3. The intermicellar interaction

To compute the DLVO interaction between micelles it isnecessary to obtain both energies of interaction between mi-celles, the electrostatic (WE(D)) and the van der Waals one(WvdW(D)).

To compute the micelle radius rM it is necessary to knowthe micelle aggregation number n and the micellized surfac-tant molecule volume. The aggregation number of SPFO wastaken as 17.5, as an average between nSPFO = 20 [21] and15 [22], and that of DTAB as nDTAB = 54 [19, p. 120]. The par-tial molar volume of SPFO at 25 ◦C is PMVSPFO = 213.00 ±0.02 cm3 mol−1 [23], and that of DTAB is PMVDTAB =278.1 cm3 mol−1 [24]. The aggregation number of mixed mi-celles was estimated as nmM = XSPFOnSPFO + XDTABnDTAB.This is a very crude estimation and as a consequence, thefollowing results must be taken only as estimations to re-veal tendencies, but not as absolute values. The volume of themixed micelle was estimated as VM = nmM(XSPFOPMVSPFO +XDTABPMVDTAB). From this value rM was estimated, and us-ing the experimental value of ζ instead of ψ0, the WE(D)

430 J.L. López-Fontán et al. / Journal of Colloid and Interface Science 312 (2007) 425–431

value was computed with Eq. (15) in Supplementary mater-ial.

The non-retarded Hammaker constant AMM of micelles wastaken as that of the hydrophobic core, since it is the majorityof the micelle volume. On the assumption of volume additivityof the Hammaker constant, AMM was computed from the val-ues for fluorocarbon and hydrocarbon. We used the values forpoly(tetrafluoroethylene) (PTFE): APTFE = 3.8 × 10−20 J, andthat of dodecane: AC12H26 = 5.0 × 10−20 J [13, p. 186] and theequation:

(6)

AMM = XSPFOPMVDTABAPTFE + XDTABPMVDTABAC12H26

XSPFOPMVDTAB + XDTABPMVDTAB.

Then, the Hammaker constant for the interaction betweentwo equal micelles immersed in water was computed withEq. (17) in Supplementary material and AWW = 3.7 × 10−20 J[25, p. 186], the Hammaker constant for pure water. Then, theattractive potential for the interaction between two micelleswhose surfaces are separated by the distance D was computedwith Eq. (14) in Supplementary material.

Fig. 3 in Supplementary material shows the dependence ofWDLVO on D for micelles having between αSPFO = 0 (pureDTAB) and αSPFO = 0.5. It may be seen that pure DTABmicelles have an energy barrier much higher than kBT , butincreasing αSPFO produces a reduction of the energy barrier,which for αSPFO = 0.3 is below 3kBT and for αSPFO = 0.5there is not energy barrier. Fig. 4 in Supplementary materialshows WDLVO vs D for the remainder systems. Because of thehigh ζ potential there are high energy barriers. However, all theWDLVO values were computed at the CMC. An increase in thesurfactant concentration may change the micelle composition,and increase the ionic strength. Both effects can reduce the en-ergy barrier and produce the flocculation of micelles giving riseto the formation of the coacervate. Comparison between Figs. 2and 9 (or Fig. 2 in Supplementary material) shows that the sys-tems having higher ζ potential at the CMC have higher valuesof C3. Micelles probably remain in clusters, because the coacer-vate is not birefringent, i.e., there is not formation of gel, crystalor any mesophase (except a cubic one). Since the coacervatewas not birefringent on flow, the flowing units (i.e., the clus-ters) could not be asymmetric. However, it must be consideredthat the computations of the DLVO energy are only estimations,because of the approximations used. Nevertheless, the generalconclusions must be correct.

5. Concluding remarks

• The interaction between SPFO and DTAB is strongly syn-ergistic.

• Above αSPFO = XSPFO = 0.38, the interaction is repulsiveand increases with the content of SPFO in both, the mixtureand micelles, whereas the interaction is attractive if DTABis in excess (XSPFO � 1 − 0.38 = 0.62). At XSPFO = 0.38the mutual low miscibility between hydrocarbon and fluo-rocarbon and the electrostatic attraction between the oppo-site charged head groups cancel each other.

• The solubility of fluorocarbon chains in the hydrocarbonmicelle core is lower than that of hydrocarbon tails in thefluorocarbon micelle core. Micelles of DTAB act as a sol-vent for SPFO without important structural changes, whilstmicelles of SPFO undergo important changes when dis-solve DTAB, and the mixture is much more nonideal thanof SPFO in DTAB micelles. This asymmetry may be in-terpreted as caused by the difference in chain length thatfavors the inclusion of the shorter chain in micelles of thelonger surfactant, but disfavors the opposite process.

• Above XSPFO = 0.5 there is an adsorption of an excess ofbromide ions on the mixed micelles, giving rise to a high ζ

potential. This adsorption is not electrostatic but originatedin van der Waals forces.

• Micelles of pure SPFO or pure DTAB show an importantenergy barrier which prevents micelle flocculation. The in-clusion of SPFO in DTAB micelles produces a reduction ofthe energy barrier, which disappeared when αSPFO = 0.5.This situation and the increase in ionic strength by increas-ing the surfactant concentration produces the flocculationof micelles giving rise to the formation of the coacervate.Since this coacervate is not birefringent, it must be formedby unordered isometric clusters of micelles.

• In summary, the above results indicate the complexity ofinteractions involved in fluorocarbon–hydrocarbon surfac-tant mixed micelles: the low mutual solubility of hydro-carbons and fluorocarbons is superimposed to the effect ofthe difference in chain length, and in the case of catan-ionic mixtures, the attraction between head groups mustalso be taken into account. The reduction in the surfacecharge causes the formation of a coacervate by agglomer-ation of micelles. Moreover, micelles having low surfacecharge adsorb bromide ions by van der Waals forces, evenon originally negative micelle surface.

Acknowledgments

This research was funded by the Spanish “Ministerio de Ed-ucación y Ciencia, Plan Nacional de Investigación (I+D+i),AT2005-02421” and from “European Regional DevelopmentFund, (ERDF)”. P.V.M. thanks Fundación Antorchas, Argen-tine, and Banco Rio, Argentine, for financial support (Project1408/110) and a grant of the Universidad Nacional del Sur, andthe Agencia Nacional de Promoción Científica y Tecnológica(ANPCyT; fund PICT No 10-14560), Argentine.

Supplementary material

The online version of this article contains additional supple-mentary material.

Please visit DOI: 10.1016/j.jcis.2007.03.035.

References

[1] P.M. Holland, Adv. Colloid Interface Sci. 26 (1986) 111.[2] W. Guo, B.M. Fung, S.D. Christian, E.K. Guzman, in: P.M. Holland, D.N.

Rubingh (Eds.), Mixed Surfactant Systems, in: ACS Symposium Series,vol. 501, Am. Chem. Soc., Washington, DC, 1992, chap. 15, pp. 244–254.

J.L. López-Fontán et al. / Journal of Colloid and Interface Science 312 (2007) 425–431 431

[3] G.X. Zhao, B.Y. Zhu, in: J.F. Scamehorn (Ed.), Phenomena in Mixed Sur-factant Systems, in: ACS Symposium Series, vol. 311, Am. Chem. Soc.,Washington, DC, 1986, p. 184.

[4] D.J. Iampietro, E.W. Kaler, Langmuir 15 (1999) 8590.[5] A.V. Barzykin, M. Almgren, Langmuir 12 (1996) 4672.[6] J.C. Eriksson, S. Ljunggren, U. Henriksson, J. Chem. Soc. Faraday Trans.

2 81 (1985) 833.[7] H. Hoffmann, G. Pössnecker, Langmuir 10 (1994) 381.[8] T. Forster, W. von Rybinski, M.J. Schwuger, Tenside Surf. Deterg. 27

(1990) 254.[9] I.W. Osborne-Lee, R.S. Schechtere, in: J.F. Scamehorn (Ed.), Phenomena

in Mixed Surfactant Systems, in: ACS Symposium Series, vol. 311, Am.Chem. Soc., Washington, DC, 1986, p. 30.

[10] M. Funasaki, in: K. Ogino, M. Abe (Eds.), Mixed Surfactant Systems,Dekker, New York, 1993, chap. 5, pp. 146–188.

[11] B.V. Derjaguin, L. Landau, Acta Physicochim. USSR 14 (1941) 633.[12] E.J.W. Verwey, J.Th.G. Overbeek, Theory of the Stability of Lyophobic

Colloids, Elsevier, Amsterdam, 1948.[13] K.L. Stellner, J.C. Amante, J.F. Scamehorn, J.H. Harwell, J. Colloid Inter-

face Sci. 123 (1988) 186.[14] K.L. Stellner, J.F. Scamehorn, Langmuir 5 (1989) 70.

[15] K.L. Stellner, J.F. Scamehorn, Langmuir 5 (1989) 77.[16] J.H. Hildebrand, J.M. Prausnitz, R.L. Scott, Regular and Related Solu-

tions, Van Nostrand–Reinhold, New York, 1970, chap. 10.[17] E. Kissa, Fluorinated Surfactants and Repellents, second ed., Dekker, New

York, 2001, p. 289.[18] P.C. Schulz, J.L. Rodríguez, R.M. Minardi, M.B. Sierra, M.A. Morini,

J. Colloid Interface Sci. 303 (2006) 264.[19] N.M. van Os, J.R. Haak, L.A.M. Rupert, Physico-Chemical Properties of

Selected Anionic, Cationic and Nonionic Surfactants, Elsevier, Amster-dam, 1993, p. 118.

[20] J.L. López-Fontán, F. Sarmiento, P.C. Schulz, Colloid Polym. Sci. 283 (8)(2005) 862.

[21] G. Sugihara, P. Mukerjee, J. Phys. Chem. 85 (1981) 1612.[22] P. Mukerjee, M. Korematsu, M. Okawauchi, G. Sugihara, J. Phys.

Chem. 89 (1985) 5308.[23] S. Milioto, R. Crisantino, R. De Lissi, A. Inglese, Langmuir 11 (1995)

718.[24] D.E. Guveliu, J.B. Kayes, S.S. Davis, J. Colloid Interface Sci. 82 (1981)

307.[25] J.N. Israelachvili, Intermolecular and Surface Forces, second ed., Acad-

emic Press, London, 1991, p. 243.