Langmuir 1994,10,3180-3187
3180
Mixed Micelles Formed by Cationic Surfactants and Anionic Hydrophobically Modified Polyelectrolytes Benoit Magny,t Ilias Iliopoulos,*BtRaoul Zana,$ and Roland Audebertt Laboratoire de Physico-Chimie Macromolbculaire, Universite Pierre et Marie Curie, CNRS URA - 278, ESPCI, 10 rue Vauquelin, 75231 Paris Cedex 05, France, a n d Institut Charles Sadron, 6 rue Boussingault, 67083 Strasbourg Cedex, France Received December 17, 1993. I n Final Form: June 9, 1994@ The association between the cationic surfactant dodecyltrimethylammonium chloride (DTAC) and a series ofhydrophobicallymodified poly(sodium acrylate) (HMPA)was investigated by rheology and by both steady-stateand time-resolved fluorescence. The polymers contain 1 or 3 mol % alkyl side groups (dodecyl or octadecyl). The strength of the interaction is improved when the polymer hydrophobicityincreases. In the presence of modified polymer, mixed micelles are formed containing surfactants and polymer alkyl groups. Depending on the polymer and surfactant concentrations,the mixed micelles contain alkyl groups belonging t o one or more polymer chains. In the later case, cross-linking between polymer chains occurs, leading to the viscosification or gelation ofthe system. Steady-state and time-resolved fluorescence methods have been used to estimate the aggregation number of the mixed micelles. It was found that the total number of surfactant molecules and alkyl groups in the mixed micelles is close to the aggregationnumber of the free surfactant micelles. From the above results the number of alkyl groups in the mixed micelles was derived and correlated to the rheological behavior of these systems. The importance of inter- and intrachain contributions to the formation of the mixed micelles is discussed.
Introduction
such systems is, in general, not symmetrical and that the concentrated phase is more often in equilibrium with a Ionic surfactants interact strongly with oppositely dilute surfactant solution than with a polymer-containing charged polyelectrolytes and form micelle-like clusters a t phase. surfactant concentration much lower than the critical The peculiar effect of hydrophobic interactions on the micelle concentration (cmc) of the ~ u r f a c t a n t . l - ~The polymer/surfactant association is clearly shown for sysconcentration a t which the surfactant begins to bind to tems containing hydrophobically modified (or associating) the polyelectrolyte is named the critical aggregation water soluble polymers (HMWSP). These are copolymers concentration (cac), and it is 1-3 orders of magnitude of a water-soluble monomer (large excess) and a very lower than the cmc. hydrophobiccomonomer (in small fraction, typically lower Although the polyelectrolyte/surfactant complex is than 5 mol %).12-23 In aqueous solution the hydrophobic stabilized mainly by electrostatic attractions, hydrophobic moieties self-associate into intra- and interchain aggregates presumably of micellar type. Because of interinteractions between the surfactant alkyl tail and polymer backbone appear to play a n important role. I t has been chain associations, the semidilute solutions of HMWSP reported that an increased hydrophobicity of the polymer exhibit much higher viscosities than the solutions of the strengthens the polyelectrolyte/surfactant i n t e r a c t i ~ n , ~ ~ ~ corresponding unmodified polymer.14-16~1s-21~23 This association, and in turn the viscosity, may be enhanced upon but a t the same time decreases the cooperativity of the addition of salt. For instance, the viscosity of an aqueous binding p r o c e ~ s .Of ~ ~course, ~ the flexibility and charge density of the polyelectrolyte chain also influence the (8)Thalberg, K.; Lindman, B.; Karlstrom, G. J . Phys. Chem. 1991, association process. 95, 3370. Another important consequence of the surfactant bind(9) Thalberg, K.; Lindman, B.; Karlstrom, G. J . Phys. Chem. 1991, 95, 6004. ing on the oppositely charged polyelectrolyte is that (10) Thalberg, &; Lindman, B.; Bergfeldt, K.Langmuir 1991,7,2893. associative phase separation takes place? a concentrated (11)Thalberg, K.; van Stam, J.;Lindblad, C.;Almgren, M.; Lindman, phase containing polyelectrolyte and surfactant is in B. J.Phys. Chem. 1991, 95, 8975. equilibrium with a dilute solution of the component in (12) Polymers i n Aqueous Media Performance through Association; Glass, J. E., Ed.; Advances in Chemistry Series, Vol 223; American excess (polymer or surfactant). In their extensive studies Chemical Society: Washington, DC, 1989. Thalberg et al.'-I1 have shown that the phase diagram of (13) Water Soluble Polymers. Synthesis, Solution Properties and
T To whom correspondence should be addressed. ' Universite Pierre e t Marie Curie.
Institut Charles Sadron. Abstract published inAduance ACSAbstracts, August 15,1994. (1)Goddard, E. D. Colloids Surf. 1986, 19, 301. (2) Hayakawa, K.; Kwak, J. C. T. In Cationic Surfactants Physical Chemistry; Rubingh, D., Holland, P. M., Eds.; Marcel Dekker: New York, 1991; p 189. (3) Lindman, B.; Thalberg, K. In Polymer-Surfactant Interactions; Goddard, E. D., Ananthapadmanabham, K. P., Eds.; CRC Press: Boca Raton, FL, 1992. (4) Shimizu, T.; Seki, M.; Kwak, J. C. T. Colloids Surf. 1986,20,289. (5) Benrraou, M.; Zana, R.;Varoqui, R.; Pefferkorn, E. J . Phys. Chem. 1992, 96, 1468. (6) Piculell, L.; Lindman, B. Adu. Colloid Interface Sci. 1992, 41, 149. (7) Thalberg, K.; Lindman, B.; Karlstrom, G. J . Phys. Chem. 1990, 94, 4289. @
0743-7463/94/2410-3180$04.50/0
Applications; Shalaby, S. W., McCormick, C. L., Butler, G. B., Eds.; ACS Symposium Series, Vol. 467; American Chemical Society: Washington, DC, 1991. (14) McCormick, C. L.; Nonaka, T.; Johnson, C. B. Polymer 1988,29, 731. (15) Schulz, D. N.; Kaladas, J. J.;Maurer, J . J.; Bock, J.;Pace, S. J.; Schulz, W. W. Polymer 1987,28, 2110. (16) Landoll, L. M. J . Polym. Sci. Polym. Chem. Ed. 1982,20, 443. (17) Valint, P. L.; Bock, J . Macromolecules 1988,21, 175. (18)Wang, T. K.; Iliopoulos, I.; Audebert, R. In ref 13, Chapter 14, p 218. (19) Biggs, S.; Hil1,A.; Selb, J.;Candau, F. J . Phys. Chem. 1992,96, 1505. (20) Biggs, S.; Selb, J.; Candau, F. Polymer 1993, 34, 580. (21)Rauscher, A,; Hoffmann, H.; Rehage, H.; F'ock,J . Tenside, Surfactants, Deterg. 1992, 29, 101. (22) Maechling-Strasser, C.; Franqois, J.; Clouet F.; Tripette, C. Polymer 1992, 33, 627. (23) Tanaka, R.; Meadows, J.; Williams, P. A,; Phillips, G. 0. Macromolecules 1992,25, 1304.
0 1994 American Chemical Society
Mixed Micelles
Langmuir, Vol. 10,No. 9,1994 3181
solution of hydrophobically modified polyelectrolyte increases with the ionic strength.l8Sz0 A more intriguing behavior was found when HMWSP were mixed with surfactants. The viscosity of such mixtures goes through a pronounced maximum as the surfactant concentration is increased at constant polymer concentration.20~z3-z* This behavior has been attributed to the formation of mixed micelles of surfactants and polymer hydrophobic group^.^^-^^ It must be noted that polyelectrolytes containing a low fraction of very hydrophobic groups, for instance octadecyl chains, can associate even with surfactants of the same charge.26 In this case the attractive hydrophobic interactions overcome the unfavorable electrostatic repulsions between the polymer backbone and the surfactant ionic heads. We have previously studied the effect of hydrophobic modification of poly(sodium acrylate) (PA) on its interactions with anionicz6and nonionic surfactants.28 In this paper we focus on the association between the same polymers and cationic surfactants (dodecyltrimethylammonium bromide (DTAB) or chloride (DTAC)). The hydrophobically modified derivatives of poly(sodium acrylate) (HMPA) bear 1or 3 mol % ofN-dodecyl(or octadecyl) acrylamide units randomly distributed along the poly(sodium acrylate) b a c k b ~ n e They . ~ ~ were ~ ~ ~obtained by reacting 1 or 3% of the carboxyl groups of a precursor poly(acry1ic acid) with N-dodecyl (or octadecyl)amine.18 The surfactant binding on the HMPA induces a very pronounced viscosity eghancement not observed with the unmodified precursor poly(sodium acrylate). The critical aggregation concentration decreases by increasing the hydrophobicity of the polymer, but at the same time, increased amounts of surfactant are needed to induce associative phase separation (precipitation). This behavior was attributed to the formation of mixed micelles between HMPA and cationic surfactants,the size of which (aggregation number) was estimated by using steady-state and time-resolved fluorescence methods. The number of side alkyl groups of the polymer in the mixed micelles was then derived and correlated to the rheological behavior of the systems.
Experimental Section Materials. The origin of poly(sodium acrylate),the modification reaction, and the polymer characterization have been described.'8.29 The molecular structure of the modified samples is the following:
c=o I ONa'
3 ,i-c12
Id 10'
10'~
lo4
lo3
10''
(mod) Figure 1. Variation of the viscosity of 1%aqueous solution of Surfactant concentration
the precursor and modified polymers with surfactant concentration: empty symbols, CTAB;full symbols,CTAC. The hatch marks represent the limit of phase separation.
hexadecylpyridinium chloride (HPC)were obtained from Fluka and purified by two crystallizationsfrom ethanollethyl acetate mixtures. Highly purified pyrene ('99.9%) was obtained from Polysciences Inc. Water was purified by a Milli-Q system (Millipore). Preparation of Polymer/Surfactant Mixtures. Polymer/ surfactant mixtures of the desired composition were prepared by mixing aqueous stock solutions of polymer and surfactant. The concentrationof the stock solutions was twice that in the final mixture. The mixtures were shaken vigorously several times and equilibrated for 1-2 days at room temperature. A microquantity of an ethanolicstock solution of pyrene (6 x M)was added to the solution of polymer/surfactantmixture to obtain the desired final concentrationin pyrene, 3 x to 1.2 x 10-6 M, depending on the experiment. Concentrated stock solutions of the quencher were prepared and added with a microsyringeto the mixtures. In the time-resolved fluorescence quenching(TRFQ)experiments,the quencher concentration[Q] was adjusted to be close to the micelle concentration[MI. For the steady-state fluorescence quenching (SSFQ)measurements, [Ql was varied from [MY3 to 3[Ml. For the TRFQ experiments the solutionswere deoxygenatedprior to use by three successive freeze-pump -thaw cycles. Apparatus. Fluorescence steady-statespectrawere recorded on a AMINCO 500 SPF spectrometer. The excitation wavelength was 334 nm, and band passes were set at 5 and 0.5 nm for excitation and emission, respectively. The emission intensities at 373 and 384 nm were used to calculatethe first to third peak intensity ratio (Z1/13). The intensity of the first peak (11) at 373 nm was used to estimatethe aggregation numbersfrom the SSFQ experiment^.^^
c=o I HN-(CH,)
5
- CH,
n-1
where x is the modification degree in mol % and n the number of carbon atoms of the alkyl chain. Here x = 1or 3 and n = 12 or 18. The modified polymers are denoted,for instance, as 1C18 for a sample modified 1mol % with octadecyl groups ( x = 1, n = 18). All samples are used in the fully neutralized sodium form: their aqueous solutionshave a pH close to 9. The degree of polymerization (= 2000) is the same for modified and unmodified samples. The surfactantsDTAB and DTAC were purchased from Tokyo Kasei as 99%pure and dried under vacuum (10-2 mmHg)before use. The quenchers dodecylpyridinium chloride (DPC) and (24) Gelman, R. A. Int. Dissolving Pulps Conf. (TAPPI Proc.) 1987, 159. (25)Biggs, S.;Selb, J.; Candau, F. Langmuir 1992,8, 838. (26)Iliopoulos, I.;Wang, T. K.; Audebert, R. Langmuir 1991,7,617. (27) Magny, B.; Iliopoulos, I.; Audebert, R.; Piculell, L.; Lindman, B. Prog. Colloid Polym. Sei. 1992,89,118. (28)Iliopoulos, I.; Olsson, U. J. Phys. Chem. 1994,98,1500. (29)Magny, B.; Lafuma, F.; Iliopoulos, I. Polymer 1992,33, 3151.
The fluorescence decay curves were determined using the single-photon counting setup described in previous s t ~ d i e s . 3 ~ ~ ~ ~ The time-decay curves were fitted to the Infelta equation33by using a nonlinear weighted least squares procedure. Most of the viscosity measurements were performed with a Contraves LS-30 viscometer at low shear rates (between 0.06 and 1.28s-1) correspondingto the Newtonian viscosity. For very viscous mixtures (v 2 100000 cp) a Carri-Med controlled stress rheometer equipped with a cone and plate geometry was used, and viscosities at a shear rate of 0.1 s-l are reported. All the measurements were performed at 25.0 & 0.1 "C. The polymer concentrationwas kept constant at 1%by weight.
Results and Discussion Rheology. Figure 1 shows the variation of the viscosity as a function of DTAB concentration for 1% aqueous solutions of PA, 1C12, and 3C12. The precursor, unmodi(30)Turro, N.J.;Yekta, A. J . Am. Chem. SOC. 1978,100,5951. (31)Binana-Limbel6, W.;Zana, R. Macromolecules 1987,20,1331. (32)Binana-Limbel6, W.; Zana, R. Macromolecules 1990,23,2731. (33)Infelta, P.; Gratzel, M.; Thomas, K. J . Phys. Chem. 1974,78, 190.
Magny et al.
3182 Langmuir, Vol. 10, No. 9, 1994 fied poly(sodium acrylate), exhibits only a very slight decrease in viscosity when the surfactant concentration C is increased. As expected, phase separation occurs (precipitation)6-*0for surfactant concentrations higher than a critical value that will be referred to as the critical precipitation concentration (cpc). Note that the phase separation occurs well before the charge stoichiometry. The molar concentration of a 1%PA solution is 0.106 M, and the corresponding cpc for DTAB is of about 3 x M. In such phase-separated systems and for surfactant concentrations close to the cpc, all the surfactant is expected to be in the precipitate while the supernatant contains practically only excess p ~ l y m e r . ~This . ~ is probably due to the highly cooperative ~ h a r a c t e r of ~.~ the binding process of the surfactant onto the polyelectrolyte chain. An even distribution of the surfactant molecules over all the polyelectrolyte chains should correspond to a noncooperative binding. A very different rheological behavior is observed when modified polymers are used instead of PA (Figure 1).The viscosity startsto increase at C above about M, passes through a pronounced maximum, and then decreases M. The viscosity maximum sharply at C higher than is highly dependent on the modification degree ( x ) of the polymer and on the length (n)of the side alkyl group: the higherx or n the more pronouncedthe viscosity maximum. The position of the maximum is not always easy to determine. Many of the modified polymers form very viscous solutions or elastic gels over a rather wide range of surfactant con~entrations,2~ generally between and M (see Figure 1, curve 3C12). Another interesting observation is that the phase separation limit occurs at much higher C for the modified polymer, about 2.5 x M, than it does with the PA (% 3x M). Obviously the surfactant molecules (or the surfactant clusters) distribute more evenly over the modified polymer chains than they do with the precursor. This is an indication that the cooperativity of the binding process of an ionic surfactant to an oppositely charged polyelectrolyte decreases when the polymer hydrophobicity increases as also reported for other polymer/surfactant system^.^*^ The viscosity enhancement and the increased cpc can be understoodif one assumes that in the presence of HMPA the cationic surfactant comicellizes with the side alkyl groups of the polymer. The four-step mechanism in Figure 2 schematically represents this process. The first part of the figure schematically depicts the polymer solution in the absence of surfactant. At this polymer concentration there is no, or only very little, self-aggregation of the side alkyl groups. The viscosity of this solution is close to that of the unmodified polymer. When the surfactant concentration is close to the cac of the system, the surfactant molecules start to bind on the polymer chain (Figure 2b). This binding occurs preferentially close to the polymer alkyl groups, and its cooperativity is probably low. At this stage the amount of the surfactant added is not sufficient to bind around all of the polymer alkyl groups. Some mixed hydrophobicclusters (presumably of micellar type) form and start to cross-link the polymer chains. When more surfactant is added (Figure 2 4 , the number ofmixed micelles increases, the number of alkyl groups per micelle decreases, but progressively all of them are included in the clusters. A highly connected tridimensional network is then formed, and the viscosity of the mixture increases or gelation may occur. Further addition of surfactant has two consequences. The excess surfactant continues to be involved in the formation of mixed micelles in which the ratio [alkyl group~surfactant]is low. In a borderline case the mixed micelles may contain only one or a few
J
-t
1
surfactant
El
i
n
\
b
B-L
micelle
Figure2. Schematicrepresentation of the association process between HMPA and a cationic surfactant: (a)HMPA solution in the absence of surfactant; (b) the same solution as in part a but in the presence of a low amount of surfactant, C < cac; (c)more surfactantis added, C > cac. Mixed micellesare formed, leading to the cross-linking of the polymer chains. (d)Further addition of surfactant results in the formation of mixed micelles with a low content of polymer alkyl groups and of micelles free of alkyl groups bound electrostatically onto the polymer backbone. The system approaches the phase separation limit. alkyl groups belonging to the same polymer chain. The cross-linking of polymer chains decreases as well as the viscosity of the system. The surfactant can also start to form clusters free of alkyl groups bound on the polymer through electrostatic interactions. These clusters are similar in nature to those formed when the unmodified anionic polymer is mixed with a cationic surfactant. After this stage, the phase separation occurs similarly as in the PA/DTAB system. According to the above mechanism, a large number of surfactant molecules are involved in the formation of mixed micelles around the polymer alkyl groups. This leads to a more even distribution of the surfactant molecules over all the polymer chains and can explain the increased cpc observed for the modified polymer. From the viscosity curves shown in Figure 1it is clear that when the surfactant counterion changes from bromide (empty symbols) to chloride (full symbols), there is no detectable change in the viscosity of the mixture. We conclude that the chloride and bromide salts of dodecyltrimethylammonium have the same affinity for the poly(sodium acrylate) derivatives. A rather similar affinity of bromide and chloride salts of decyltrimethylammonium toward a maleic acid copolymer has been reported by Binana-Limbel6 e t al.31 In the fluorescence experiments we prefer to use the chloride salt to avoid undesirable quenching of pyrene by bromide anions. Critical Aggregation Concentration. The mechanism illustrated by Figure 2 is based on the assumption that the affinity of the surfactant for the polymer chain is increased when the polymer bears hydrophobic side chains because both electrostatic and hydrophobic inter-
Mixed Micelles
Langmuir, Vol. 10, No. 9, 1994 3183 DTAC in 0.1 M NaCl
I
Table 1 sample no polymer 0.1 M NaCl
PA 1c12 3C12 1C18
1.4
1
I-".
I
t 105
104
10"
10'2
10''
DTAC concentration(mol P)
Figure 3. 11/13 as a function of DTAC concentration in the absence of polymer in water and in 0.1M NaC1, and in the presence of 1% polymer (PA,lC12,3C12,1C18) in water. Pyrene concentration = 6 x lo-' M. actions are operative. The cac of the surfactant in such a system is expected to be lower than the corresponding one found with the unmodified poly(sodium acrylate) system. The cac can be obtained by using the steady-state fluorescencemethod with pyrene as a fluorescenceprobe.% The fluorescence emission spectrum of pyrene shows several vibronic peaks the intensity of which is very sensitive to the polarity of its environment. Especially, the ratio of the intensities of the first and third vibronic peaks (11113)is a sensitive indicator of the polarity of the pyrene m i c r ~ e n v i r o n m e n t .If~ ~micelles ~ ~ ~ or other hydrophobic microdomains are formed, in an aqueous solution, the pyrene preferably lies close to (or inside) these microdomains. The formation of micelles or hydrophobic aggregates can then be followed by measuring the ratio 11/13.34-41For instance, 11/13varies from 1.85in pure water to 1.14 in a micellar solution of SDS.34 This technique was successfully used to determine the cac in polymerlsurfactant s y ~ t e m s . ~ l > ~ ~ - ~ ~ In Figure 3, 11/13is plotted as a function of the DTAC concentration in the presence of 1% polymer (PA, 1C12, 3C12, and 1C18). For comparison are given the curves in the absence of polymer, in water, and 0.1 M NaC1. The curve corresponding to DTAC in pure water can be used to determine the cmc of this surfactant. The sharp transition in 11/13 a t surfactant concentrations between 1.5 x and 2 x M indicates the formation of micelles. The cmc corresponds to the end ofthis transition (2 x lov2 and it is in very good agreement with the reported value.42 In the presence of 1% unmodified PA the transition shifts to much lower (20 times) surfactant concentrations. This indicates the formation of surfactant clusters bound onto the polyelectrolyte, and the end ofthe (34)Kalyanasundaram,K.;Thomas, J. K. J.Am. Chem. SOC.1977, 99,2039. (35)Malliaris, A. Znt. Rev. Phys. Chem. 1988,7,95. (36)Frindi, M.;Michels, B.;Zana, R. J . Phys. Chem. 1991,95,4832; 1992,96,6095;1992,96,8137. (37)Chu, D.Y.;Thomas, J. K. J . Am. Chem. SOC.1986,108,6270. (38)Turro, N.J.;Baretz, B. H.; Kuo, P. L.Macromolecules 1984,17, 1321. (39)Chandar,P.;Somasundaram,P.;Thomas,J. K. Macromolecules 1988,21,950. (40)Winnik, F . M.; Winnik, M. A.; Tasuke, S. J.Phys. Chem. 1987, 91,594. (41)Winnik, F. M.; Ringsdorf, H.; Venzmer, J. Langmuir 1991,7 , 905. (42)Lindman, B.;Wennerstrom, H. In Topics in Current Chemistry; Boschke, F.L., Ed.; Springer-Verlag,New York, 1980;Vol. 87,p 1.
cmc or cac (M) 2 x 10-2 6x 1 x 10-3 6x 2 10-4 5 x 10-5
transition at low3M DTAC is taken as a measure of the cac of the system.37 The shape ofthe curve in the presence and in the absence of PA is practically the same, indicating similar cooperativities for micelle formation in the two systems. The lower value of the cac as compared to the cmc in the absence of PA may be partially due to the increased ionic strength of the solution induced by the polyelectrolyte. A 1%by weight solution of PA has a molar concentration in repeating ionic units of about 0.1 M. When 0.1 M NaCl solution is used as solvent, the cmc of DTAC is x6 x M (Figure 3). This value is still 6 times higher than the cac in the presence of PA. The introduction of hydrophobic groups onto the polyelectrolyte chain greatly influences the shape of the curve and decreases the cac (Figure 3). The decrease of 11/13 becomes less steep but remains much steeper than what is expected for a partition of pyrene between aggregates and water. Indeed, in such a case the decrease of 11/13 would stretch over a factor of more than lo2in concentrat i ~ n . The ~ ~ pyrene ? ~ senses a more gradual increase in the hydrophobicity of the microdomains than it does with PA. This more gradual change of 11/13 with C suggests that the polymerlsurfactant association process is less cooperative when the polymer hydrophobicity increases and indicates that the structure of the aggregates formed between CTAC and HMPA differs from that of the CTACI PA aggregates. Presumably there is formation of premicellar-type aggregates in which the polymer alkyl groups are involved as proposed in Figure 2b. The structure of these aggregates changes gradually with increasing C . Besides, the results in Figure 3 clearly show that the affinity of the surfactant for the oppositely charged polyelectrolyte drastically increases with the polymer hydrophobicity. The cac decreases when the alkyl chain length or the degree of modification of the polymer increases (Table 1). Similar conclusions about increased affinity, lower cooperativity, and preferential binding of the surfactant close to the hydrophobic parts of the polyelectrolyte, a t least during the first steps of the binding, were reported for the association between cationic surfactants and alternating copolymers of maleic acid with various comonom e r ~ . ~More , ~ specifically, a systematic comparison between the binding isotherms and the variation of 11/13 with C showed that, in practice, a decrease in the cooperativity of the polymerlsurfactant association process is coupled with a smoothness of the transition in the 11/13 vs C curve.5 Another interesting result seen in Figure 3 is the plateau value of 11/13before the transition (C versus EQI were found for EQl/[Ml formation of mixed micelles and eq 2 becomes 5 2 (Figure 4). The most hydrophobic HPC is generally preferred as quencher to DPC when used in cationic Nt = (C, C - cac)/[Ml (3) surfactant systems to avoid intermicellar exchange of the q u e n ~ h e r .In ~ ~our systems we checked that both DPC where Nt is the total aggregation number of the mixed and HPC can be used. Table 2 shows that either in a micelle (alkyl groups surfactants) and C , the alkyl group DTAC solution or in a DTAC/3C12 system the same molar concentration. Taking into account that the aggregation numbers were obtained independently of the polymer concentration in the mixtures is 1%, the total quencher used. In this study, DPC was preferred to HPC molar concentration of alkyl groups is found to be about because the more hydrophobic HPC was suspected to M for 1C12 and 1C18 and 3 x M for 3C12. The exhibit, in some cases, very strong interactions with HMPA (much stronger than the surfactant DTAC). (45) Zana, R. In Surfactant Solutions New Methods of Investigation; The total aggregation numbers of the mixed micelle Nt Zana, R., Ed.; Surfactant Sciences Series, Vol. 22; Marcel Dekker: New and the numbers of alkyl groups per micelle N , for the York, 1987; Chapter 5. various systems studied are listed in Table 3. The (46) Zana, R.; Lianas, P.; Lang, J. J . Phys. Chem. 1986, 89, 41. (47) Binana-Limbel6,W.; Zana, R. Colloids Surf. 1986,21, 483. aggregation number of pure DTAC micelles (N, = 42) is (48) Zana, R.; Binana-LimbelB, W.; Kamenka, N.; Lindman, B. J. found to be lower than that reported previously ( N ,= 54) Phys. Chem. 1992, 96, 5461. by Malliaris et aL51 This value (N, = 42) was also found (49) van Stam, J.;Almgren, M.; Lindblad, C. Progr. Colloid Polym.
+
+
+
Sci. 1991, 84,13. (50)Almgren, M.; Hansson, P.; Mukhtar, E.; van Stam, J. Langmuir 1992, 8, 2405.
~~~~~
(51)Malliaris, A.; Boens, N.; Luo, H.; van der Auweraer, M.; de Schryver, F. C . ; Reekmans, S. Chem. Phys. Lett. 1989, 155, 587.
Langmuir, Vol. 10, No. 9, 1994 3185
Mixed Micelles Table 4
no polymer
[DTAC] (mol L-I) 0.105
PA
2.90
sample
DPCI
(mol L-l) 0 1.53 x 10-3
10-3
o
2.55 5.30 1c12
1.70 x 10-3
o
10-5 10-5
4.80 x 10-5 3.30
10-3
o
7.95 1.84
10-5 10-4
1.30
10-4
10-3
o
1.00 x 10-2
0
5.30
1.66 2.72 3C12
1.00 x 10-2 2.00 x 10-2
10-4 10-4
0
1.66 x 10-4 0 4.80 10-4 8.18 x 10-4
fluorescence
quenching
decay const, 10+k 3.03 3.37 2.99 3.25 3.17 3.22 3.42 3.27 3.42 3.40 3.15 3.37 3.06 3.26 3.16 2.91 3.06 3.36 3.29 3.31
rate const, 10-7k,
N,orNt
N,
kdk
(TRFQ)
(TRFQ)
0.75
3.56
10.6
42
0.81 1.73
1.78 1.73
5.47 5.45
49 57
1.68
1.61
4.71
61
29
1.53 3.01
1.80 1.70
5.26 5.00
64 58
17
1.62
1.74
5.16
65
12
1.60 2.30
1.59 1.69
4.88 5.35
83 77
9
1.05
1.53
5.00
61
14
1.58 2.59
1.52 1.44
4.62 4.35
62 63
8
t (ns)
330 297 334 308 315 310 292 306 292 294 317 297 327 307 316 344 327 298 304 302
by the TRFQ experiments (see next section). When a modified polymer is added, the total aggregation number Nt is slightly increased as compared to N , of pure surfactant. The following comments can be made. For a given polymer, Nt does not change significantly, within the experimental error, when C is increased (Nt % 51 for 1C12 and 1C18 andNt % 45 for 3C12). As a consequence, the number of alkyl groups in the mixed micelles N , decreases upon increasing C, as also suggested in the model of Figure 2. However, N , never reaches the borderline value N , = 1 for the mixtures with the highest surfactant concentration (mixtures of low viscosity just before the phase separation; see Figure 1). This point is further discussed below. Note that a lower aggregation number was found for micelles bound on the precursor (N, = 30). This surprisingly low value was confirmedby repeating the experiment twice. For the polymer modified with octadecyl side chains (lCl8)no measurements were possible a t C < M due to the formation of stiff gels upon addition of DTAC. Time-ResolvedFluorescence Quenching (TRFQ). In the TRFQ method the system is illuminated by light flashes and the decay curve of the emitted fluorescence is determined (variation of the fluorescence intensity Z ( t ) with time t). In a micellar system containing a waterinsoluble probe and a quencher such that [PY[Ml 1condition is not fulfilled; see Table 4). A consequence of the above reported difference in the Nt values is that Na values were also found to be somewhat larger when measured by TRFQ (compareresults in Tables 3 and 4). Table 4 shows that the k, values found with polymer-free micelles are about twice as large as with micelles bound onto the precursor or modified polymer. Qualitatively similar results have been reported in other studies of surfactant interaction with and chargedl1s5O polymers. It appears that the polymer/ surfactant interaction always slows down the intramicellar quenching process. Recall that k, decreases when the microviscosity of the system increases.45 Such a n effect may be operative in our system and deserves further investigation. An alternative or complementary explanation is the existence of some interaction between pyrene and water-soluble polymers.46
Relation between Rheological Behaviorand Number of Alkyl Groups in the Mixed Micelles. Figure 6 correlates the viscosity of lC12DTAC and 3C12/DTAC
Figure 7. Schematic representation of intrachain basic and interchain cross-linking mixed aggregates. mixtures to the correspondingnumber of alkyl groups per mixed micelle measured by SSFQ. (The use of N,values obtained by TRFQ leads to the same qualitative conclusion.) In the first part of this paper we suggested that the decrease in viscosity at high surfactant concentration is due to the breakdown of the intermolecular aggregates (no connectivity between the polymer chains). Consequently, the mixed micelles formed when the viscosity reaches its lowest value just before the phase separation (at the right side of the viscosity peak) can be considered as a kind of basic intrachain aggregates. They contain alkyl groups all belonging to only one polymer chain. They do not produce any cross-linking of the polymer chains and, therefore, do not contribute significantly to the enhancement of the viscosity of the system. From the Na values listed in Tables 3 and 4, it turns out that these basic aggregates contain more than one polymer alkyl group. In other terms, each polymer chain contributes a number of alkyl groups (Na,basic > 1)to the formation of a given mixed micelle. Confirmation of such an extensive intrachain contribution was obtained by using a pyrenelabeled associating PA and monitoring the excimer formation.52 For the systems studied in this work, the SSFQ results indicate that Na,basic = 6 for 1C12 and 3C12 andN,, basic = 3 for 1C18. These values must not be viewed as a “fixed”stoichiometryof the mixed aggregates in alkyl groups asNavaries continuously with C. However, when C decreases to values such that Na =- Na, basic, the mixed intrachain aggregates would contain too many polymer alkyl groups to be stable, and mixed interchain aggregates start to form, inducing the increase of the viscosity. From the point of view of polymer network formation, one can viewNa,basicas the average number of alkyl groups that a polymer chain contributes to the formation of a cross-linking micelle. The viscosity maximum corresponds to the highest cross-linking of the polymer chains, that is when approximately two basic (intrachain) aggregates merge into one cross-linking (interchain) aggregate. Therefore, at the viscosity maximum the number of alkyl groups in the mixed micelles must be about 2 times that found in the basic aggregates. This is rather well confirmed by the results shown in Figure 6. A schematic representation of the basic and cross-linking aggregates is given in Figure 7. We then concludethat the increased viscosity observed for mixtures of cationic surfactant and hydrophobically (52) Magny, B. Thesis, University Pierreet Marie Curie,Paris, 1992.
Langmuir, Vol. 10,No.9,1994 3187
Mixed Micelles modified anionic polyelectrolyte is due to the formation of mixed micellar-type aggregates. However, intrachain contribution to the formation of these mixed micelles is important even in mixtures exhibiting the maximum viscosity. Note that extensive intrachain contribution was also found in mixed micelles formed between 3C18 and a nonionic surfactant C I Z E ~ . ~ ~
Conclusion Hydrophobically modified poly(sodium acrylate) associates strongly with cationic surfactants and forms mixed micelles which contain surfactant molecules and alkyl side groups of the polymer. The polymer/surfactant association starts a t surfactant concentrations well below the critical aggregation concentration observed in the presence of the unmodified PA. The cooperativity of the binding process seems to decrease upon increasing the polymer hydrophobicity. The main information of this study concerns the composition of the mixed micelles. The total aggregation
number and the number of polymer alkyl groups per micelle were determined from steady-state and timeresolved fluorescence measurements. The number of alkyl groups per micelle was found to depend on the composition of the polymer/surfactant mixture, and it was clearly shown that there is a n extensive intrachain contribution of the alkyl groups to the mixed micelles. The mixed micelles formed in the presence of a n excess of surfactant (basic aggregates), when the system loses its highly thickening properties, contain more than one alkyl group. At last, it was shown that the viscosity maximum corresponds to the formation of cross-linking mixed micelles which contain on a n average basis twice as many alkyl groups as the basic intrachain aggregates.
Acknowledgment. We thank L. Piculell, B. Lindman, A. Blanc, and G. Mattioda for many helpful discussions and K. Loyen for help in the experimental part. This work was supported by grants from the “SociBtBFranGaise Hoechst” and CNRS - France.