Fluorescence probing studies of surfactant aggregation in aqueous solutions of mixed ionic micelles

Fluorescence Probing Studies of Surfactant Aggregation in Aqueous Solutions of Mixed Ionic Micelles ANGELOS MALLIARIS, 1 WILLIAM BINANA-LIMBELE, 1 AND RAOUL ZANA 2 Institut Charles Sadron (CRM-EAHP), CNRS-ULP Strasbourg, GRECO Micro~mulsions, 6, rue Boussingault 67083 Strasbourg C~dex, France Received April 23, 1985; accepted July 17, 1985 Mixed micelles of dodecyl- and hexadecyltrimethylammonium chloride (DTAC and CTAC), and of DTAC and sodium dodecyl sulfate (SDS) have been investigated by means of conductivity and fluorescence probing. The CMC, micelle composition and mean average aggregation number of mixed DTAC-CTAC micelles have been determined. As the total surfactant concentration C is increased, the aggregation number decreases rapidly at low (7, then levels off. The decrease of Nis associated with a decrease of the CTAC content of the mixed micelles. These results illustrate the effect of a long chain surfactant impurity on the micellar properties of a shorter chain surfactant. The DTAC-SDS system shows a moderate increase ofNwhen SDS is added to DTAC and a large increase when DTAC is added to SDS. © 1986 Academic Press, Inc. INTRODUCTION Micelles formed from a mixture of two or more surfactants are of considerable interest from both practical and fundamental view- points. Thus the surfactants used for practical purposes are in most instances mixtures of surfactants differing by their chain length and/ or isomeric content (1). It is then important to know how surfactants aggregate in such mixtures and to characterize the mixed aggre- gates by their composition and aggregation number which will determine to a large extent their properties. On the other hand from the purely fundamental point of view, the study of mixed micelles contributes to the under- standing of the process of micellization (2). Thus far, most studies on mixed micelles have focused on the critical micellization con- centration (CMC) of the systems, and on the composition and ionization degree of the mixed micelles (3-15). To our knowledge there has been no determination of the aggre- gation number of mixed ionic micelles as a On leave from NRC "Demokritos," Athens, Greece. 2 To whom correspondence should be addressed. 0021-9797/86 $3.00 Copyright © 1986 by Academic Press, Inc. All fights of reproduction in any form reserved. function of the surfactant concentration. This situation stems mainly from the fact that such methods as elastic or quasielastic light scat- tering, centrifugation, osmometry, pulsed gra- dient Fourier transform NMR, etc., used for the determination of the micelle molecular weight, radius, or aggregation number require extrapolation of the data to the CMC in order to eliminate intermicellar interactions. Such an extrapolation cannot be performed with mixed micellar solutions because both the mi- celle composition (3-5, 12) and aggregation number (vide infra) depend on concentration (16). This led us to use a fluorescence probing method where the analysis of the fluorescence decay curves of a micelle-solubilized probe di- rectly yields the micelle number average ag- gregation number N under the actual experi- mental conditions, without the need of ex- trapolation (17). This method which is completely insensitive to intermicellar inter- actions, has been already extensively used for the study of micellar solutions (18) and mi- croemulsions (19). This paper reports on the aggregational be- havior of the mixed miceUes formed by mixing 114 Journal of Colloid and Interface Science, Vol. 110, No. 1, March 1986 SURFACTANT AGGREGATION 115 the cationic surfactant dodecyltrimethylam- monium chloride (DTAC) with the cationic surfactant cetyltrimethylammonium chloride (CTAC) as well as with the anionic surfactant sodium dodecyl sulfate (SDS). The DTAC- CTAC system has been investigated as a func- tion of both the mixture composition and total surfactant concentration, whereas the DTAC- SDS system was only investigated at a given surfactant concentration, as a function of composition. The three surfactants DTAC, CTAC, and SDS were selected because (i) their properties in the pure micellar solutions are well characterized (20), (ii) the results obtained with their mixtures are probably representative of what would be found with other mixtures of surfactants of like charge and opposite charge, and (iii) at low mole fraction of CTAC, the study of the DTAC-CTAC system should yield information on the effect of a long chain surfactant, considered as an impurity, on the aggregational behavior of a short chain sur- factant. The CMCs of mixtures of dodecyl and ce- tyltrimethylammonium bromides (DTAB and CTAB, respectively) were also measured as a function of the composition for the purpose of comparison with the DTAC-CTAC system. MATERIALS AND METHODS The samples of DTAC, CTAC, and SDS were the same as in previous investigations (18-20). The electrical conductivity method was used to determine the CMC of various DTAC-CTAC mixtures in the absence of py- rene, at increasing mole fraction X of CTAC. The CMC values were obtained from the plots of the conductivity r vs total surfactant con- centration C in the mixture (CMCn) and from the equivalent conductivity A vs C ~/2 plots (CMCI ) . As in other studies pyrene was used as flu- orescence probe. The fluorescence decay curves of micelle-solubilized pyrene were ob- tained using the same single photon counting apparatus as in previous studies (18-20). In the case where the pyrene distribution among micelles can be considered as "frozen" on the fluorescence timescale (i.e., when the fluorescence lifetime ofpyrene is much smaller than the time required for pyrene to migrate from micelle to micelle), the analysis of the decay curves directly yields the molar concen- tration ratio R = [pyrene]/[micelle], the pyrene fluorescence lifetime r, and the rate constant kE for intramicellar excimer formation (17). The pyrene concentration was adjusted for 0.5 < R < 1 and the number average aggregation number N of the mixed micelles was then ob- tained from N = R(C - Cc, x)/[pyrene] [1] where C is the sum of the concentrations of the two surfactants in the mixture (total con- centration), and Cc, x, the total concentration of free surfactants at total concentration C and mixture composition X (X = mole fraction of the surfactant added to DTAC). To check whether the pyrene distribution can indeed be considered as frozen we have measured for each of the investigated systems the pyrene fluorescence lifetimes r at R ~ 0.5- 1, and r0 at very low [pyrene], where R --- 0.01 and no excimer can be formed. Recall that the finding r - r0 indicates a frozen distribution, whereas r < r0 reveals that pyrene migration from micelle to micelle can occur during the lifetime of the pyrene fluorescence (21-23). When such is the case the analysis of the flu- orescence decay curves yields kE and R, and thus N, only if the mechanism for pyrene mi- gration is known. All measurements were per- formed at 25 ° . RESULTS AND DISCUSSION 1. CMC and Composition of Mixed DTAC-CTAC Micelles At low X the K "¢s C curves show a marked curvature at C > CMCH. The A vs C ~/2 plots show two distinct breaks which correspond to the onset of micellization (CMC0 (3) and, at higher C, to the apparition of nearly pure DTAC micelles. Indeed this second break oc- curs at a concentration only slightly below the Journal of Colloid and Interface Science, Vol. 110, No. 1, March 1986 116 MALLIARIS, BINANA-LIMBELE, AND ZANA CMC for pure DTAC. An identical behavior was found for the DTAB-CTAB. For the pure DTAC, CTAC, DTAB, and CTAB, the CMCx and CMCH values were found to obey the relationship CMCII -- (1.047 _+ 0.02) CMCI. A similar relationship was first reported by Mysels and Otter (3) for SDS and sodium decylsulfate (SDeS). The CMCn values for the DTAC-CTAC and DTAB-CTAB mixtures were calculated using this relation- ship. Figure 1 shows the changes of the calculated values of CMCn with X. This plot is very sim- ilar to previously reported ones (3) and to that for the DTAB-CTAB mixtures also shown in Fig. 1. The results have been fitted to Shinoda's equation (2) for the CMC of mixed micelles, using Kg as an adjustable parameter. The best fits to the data were obtained with Kg = 0.40 for DTAC-CTAC and Kg = 0.30 for DTAB- CTAB. The value of Kg for alkylcarboxylates I02[MC~'CM) 2, [ I - 4 1.5 1.0 0.5 0 ~ I i t X 0.2 0.4 0.6 0.8 FIG. 1. Variation of the calculated value of CMCII with X for the DTAC-CTAC mixture (curve 1), and DTAB- CTAB mixture (curve 2). The points on the diagonal rep- resent the Cc, x vs XM data obtained as explained in the text. The symbols used are the same as in Fig. 2. The curves 1 and 2 have been calculated as explained in the text: and sulfates was found to be around 0.56 (2). Recall that Kg is close to the micelle ionization degree a. Thus, the lower Kg found in this work for the bromide containing systems with re- spect to the chloride containing systems simply reflects the smaller micelle ionization of the former with respect to the latter systems. For example the a values of DTAC and DTAB were found to be 0.35 and 0.27 as part of this work. For each system characterized by the values of the overall CTAC mole fraction Xand total surfactant concentration C the values of Cc, x and of the CTAC mole fraction XM in the mixed micelles were determined by means of the method of Mysels and Otter (3) (see Table I). 3 Cc, x is plotted against ArM in Fig. 1. It is seen that within the experimental error the representative points fall on the diagonal which connects the representative points for pure DTAC and CTAC. The same result was found for the DTAB-CTAB mixtures. A sim- ilar observation was made by Mysels and Otter for the SDeS-SDS mixtures (3). 2. Fluorescence Probing Study of the DTAC-CTAC Mixtures: Aggregation Number of the Mixed Micelles For each of the investigated systems the val- ues of the pyrene lifetime at R "~ 0.01 and R 0.5-1 were found to be equal. This indicates that the pyrene distribution can be considered as frozen during times much longer than r0. The analysis of the decay curve then allowed the determination of R = [pyrene]/[micelle], while the micelle aggregation number N was calculated from Eq. [1 ]. The amount of mi- cellized surfactant C-Cc, x on the other hand was obtained from the K vs C curve, using the method proposed by Mysels and Otter (3). The values of N are plotted in Fig. 2 against C. The N vs C curves for X = 0.0221, 0.0378, and 0.0635 nearly coincide. They show a very rapid decrease of N upon increasing C at low C, then nearly level off at C > 0.02 M, up to 3 The reader must be warned that in Ref. (3) Figs. 5 and 6 are sometimes confused in the text due to printing errors. Journal of Colloid and Interface Science, Vol. 110, No. 1, March 1986 SURFACTANT AGGREGATION 1 17 N I I I J 100 ~ ~x-------- 1 80 70 6O 5O 40 " " "~ + / C(N/I} ] 0105 01 0.15 FIG. 2. Variation of the mixed micelle aggregation number N with the total surfactant concentration, at var- ious mole fractions of CTAC: (O) X = 0; (+) 0.0221; (@) 0.0378; (©) 0.0635; (A) 0.0928; ([3) 0.2086; (V) 0.4915; (×) 0.8768; and (tO) 1.00. C = 0.15 M. The values of N for the mixed micelles in this range coincide, within the ex- perimental error, with the value of N for pure DTAC micelles. The high values found for N at low C, reflect the fact that the micelles formed in the mixture at C close to the mix- ture's CMC, have a mole fraction of long chain surfactant much larger than the stoichiometric composition (see Table I). These results show how efficiently a small amount (1-2%) of a long chain surfactant can perturb the prop- erties (CMC, N) of a short chain homolog close to the CMC. The N vs C curves at higher X show the same trend as the ones at low X, but the initial decrease of N at low C is less steep. The curves level off for N values intermediate between those for pure DTAC and CTAC. Finally at high X, ca. above 0.8, the values of N at a given C are, within the experimental error, equal to that for pure CTAC micelles. Figure 3 shows the plot of N versus the mi- celle composition ArM for the various mixtures investigated. A sigmoidal curve with an ex- tended linear part can be drawn through all the data. With few exceptions, the scatter of experimental N values about the curve shown is within the estimated experimental error of the individual points (+15% at low C-Cx, c, +10% at high C-Cx, c on both Xr~ and N). Going back to Fig. 2 it is seen that the range where Nis nearly constant corresponds to that where XM is also nearly constant. TABLE I Experimental Data for DTAC-CTAC Mixed Micelles 102 102 C Cc~r x (M) (m) XM N 0 15.1 2.32 0 52 10.0 2.32 0 49 5.0 2.32 0 48 3.07 2.32 0 51 0.0221 15.0 2.18 0.026 53 7.5 2.18 0.036 46 1.91 1.77 0.218 51 1.83 1.71 0.228 68 0.0378 12.9 2.12 0.045 52 4.83 2.12 0.066 48 3.87 2.12 0.086 46 2.34 1.935 0.199 46 1.94 1.725 0.285 52 1.6 1.50 0.424 68 0.0635 10.47 2.0 0.078 47 7.85 2.0 0.084 49 5.20 2.0 0.101 47 1.31 1.2 0.515 72 0.0928 8.64 1.96 0.119 55 4.3 1.89 0.162 55 2.9 1.83 0.242 54 2.16 1.63 0.351 68 1.74 1.41 0.431 69 1.1 0.985 0.618 84 0.2086 11.3 1.69 0.244 64 8.0 1.70 0.263 65 5.64 1.60 0.288 64 4.8 1.50 0.304 70 2.4 1.36 0.483 84 1.2 0.89 0.720 91 0.4915 17.35 0.76 0.512 93 8.60 0.76 0.534 93 2.20 0.73 0.702 102 0.8768 10.35 0.16 0.88 104 4.4 0.16 0.88 105 1 8 0.136 1 104 4 0.136 1 102 1.25 0.136 1 102 Journal of Colloid and Interface Science, Vol. 110, No. 1, March 1986 | 18 MALLIARIS, BINANA-LIMBELE, AND ZANA I i i N 100 , , n r ~ gO o 7O 6O 50 ~'~ • .I- ~-~ - • 40 XM l 0.15 I 0.25 0.75 1.0 FIG. 3. Variation of the mixed micelle aggregation number with the CTAC mole fraction Xr~ in the mixed micelles. Same symbols as in Fig. 2. It should be noticed that N depends on the total concentration C through the changes of XM with C (rapid change of N at low C) and also because of the increase of intermicellar interactions with C (small increase of Nin the whole range of C, as shown by DTAC or CTAC). The latter introduces an additional scatter in the N vs Xr~ plot. Indeed, all the data should be taken at the same micellar concen- tration to be strictly comparable but this is not the case in Fig. 3. However the increase of N with C for the pure DTAC and CTAC micelles is only of the order of 10% in the C range investigated (see Fig. 2). Thus, to a good ap- proximation, it may be stated that in mixed micellar systems the micelle aggregation number is essentially determined by the mi- celle composition, at least at low overall sur- factant concentration. 3. Aggregation Number of Mixed DTAC-SDS Micelles The measurements were performed at a fairly high and nearly constant total surfactant concentration (C ~ 0.3-0.35 M) in order that the CMC of the mixture could be neglected in the calculations of N and that the micelle composition was close to the stoichiometric composition (see the results for the DTAC- CTAC system). The measurements were per- formed only in the SDS mole fraction range from 0 to 0.22 and from 0.86 to 1, because precipitation of dodecyltrimethylammonium dodecyl sulfate occurred in the 0.22 < X < 0.86 range, This observation is in agreement with that of Hoyer and Doerr (24) who have reported that flocculation slowly occurs in aqueous solutions of decyltrimethylammo- nium dodecylsulfate. The results are shown in Fig. 4. Both the addition of SDS to DTAC, and of DTAC to SDS result to a significant increase of N, which is much more pronounced in the case of the addition of DTAC to SDS. Recall that very large aggregation numbers have been mea- sured for decyltrimethylammonium dodecyl- sulfate micelles, at low concentration (24). Clearly these changes are associated with the neutralization of part of the micelle charge by the added surfactant. Thus the ionization de- gree of DTAC micelles as inferred from con- ductivity measurements is 0.27. If we accept that upon SDS addition there is no release of C1- counterions, the ionization degree would decrease to 0.1 upon addition of 17% SDS with respect to DTAC. Likewise an addition of 15% DTAC to SDS micelles (initial ionization de- 400 N' ' ' / ' 300 200 I x , , , , I / , , 0 0.1 0.2 0.9 FIG. 4. Variation of the mixed micelle aggregation number with the SDS mole fraction 3I. In these experi- ments the total surfactant concentration ranged between 0.3 and 0.35 M. Journal of Colloid and Interface Science, Vol. 110, No. 1, March 1986 SURFACTANT AGGREGATION 119 gree 0.35) would reduce the ionization degree to 0.2. Even though these evaluated ionization degrees are lower bound values because some counterions are probably released upon ad- dition of the oppositely charged surfactants, it remains that the SDS-rich mixed micelles (X = 0.85) are more ionized than the DTAC-rich mixed micelles (X = 0.17). Thus the micelle ionization does not appear to be the only pa- rameter which determines the changes of N upon mixed micelle formation. The packing of the chains in the micelle interior, and of the head groups at the micelle surface probably contribute to these changes. In the range 0 < X < 0.22, the values of the pyrene lifetime at low and high R values were found to be the same. On the contrary in the SDS side, the values of r0 and r were found to differ significantly (355 ns against 300 ns) for X = 0.865 which corresponds to the largest N value, nearly 400. Thus, in this system py- rene is redistributed between the mixed mi- celles, in a time comparable to to. The value of N for this system has been calculated using the full equations (21-23) for the fitting of the fluorescence decay curve. The present results give some indication concerning the parameters which determine the kinetics of probe redistribution. Recall that in ionic micellar solutions probe redistribution most likely occurs through the detachment of micelle fragments from full micelles (with a first-order rate constant k-), and the subse- quent attachment of these fragments to full micelles (20). This process will result in the transfer of probes from micelle to micelle if some of the fragments contained a probe. The comparison of the results for the DTAC-rich range (low X) and the SDS-rich range (high X) clearly shows that the parameter which de- termines k- is not the micelle charge but rather the micelle size. Indeed, the micelle charge is reduced both at low X upon addition of SDS to DTAC and at high X upon addition of DTAC to SDS in about the same manner (see above). However, the size is increased much only in the SDS-rich range and at N > 300 probe redistribution is observed to occur. The importance of the micelle size (or N) in giving rise to probe redistribution was previously pointed out for a variety of aqueous ionic mi- cellar systems (20). For the mixed surfactant systems also it appears that it is only when N is large enough, that is when the system con- tains strongly polydisperse anisodiametric mi- celles, that probe redistribution can take place. CONCLUSIONS The present study has provided the first val- ues of the aggregation number of mixed mi- celles made of DTAC + CTAC and DTAC + SDS. For the first system the N values de- pend mostly on the micelle composition at not too high total surfactant concentration. For the second system a very large increase of N has been evidenced upon addition of DTAC to SDS. The results for the system with the largest N value suggest that probe redistribu- tion takes place through micelle fragmenta- tion. ACKNOWLEDGMENTS The authors thank the PIRSEM (CNRS) for its financial support (AIP 2004). One of the authors (A.M.) thanks the French Minist~re de l'Industrie et de la Recherche for fi- nancing his stay in Strasbourg. REFERENCES 1. Cahn, A., and Lynn, J., in "Encylopedia of Technical Technology" (Kirk and Othmer, Eds.), Vol. 22, p. 332. Wiley-Interscience, New York, 1983. 2. Shinoda, K., Nakagawa, T., Tamamushi, B., and Isemura, T., "Colloidal Surfactants," Ch. 1. Aca- demic Press, New York, 1963. 3. Mysels, K., and Otter, R., J. Colloid Sci. 16, 462 (1961). 4. Kaler, E., Puig, J., and Miller, W., J. Phys. Chem. 88, 2887 (1984). 5. Calfords, J., and Stilbs, P., J. Phys. Chem. 88, 4410 (1984). 6. Rathman, J., and Scarnehorn, J., J. Phys.Chem. 88, 5807 (1984). 7. Holland, P., and Rubingh, D., J. Phys. Chem. 87, 1984 (1983). 8. Meguro, K., Tabata, Y., Fujimoto, N., and Esumi, K., Bull. Chem. Soc. Japan 56, 627 (1983). Journal of Colloid and Interface Science, Vol. 110, No. 1, March 1986 120 MALLIARIS, BINANA-LIMBELE, AND ZANA 9. Kamrath, R., and Franses, E., J. Phys. Chem. 88, 1642 (1984). 10. Nilsson, P. G., and Lindman, B., J. Phys. Chem. 88, 5391 (1984). 11. Koshinuma, M., Bull. Chem. Soc. Japan 56, 2341 (1983). 12. Lake, M., MakromoL Chem. 183, 459 (1982). 13. Kurzendorfer, C.-P., Schwuger, M.-J., and Lange, H., Bet Bunsenges. Phys. Chem. 82, 962 (1978). 14. Shinoda, K., and Nomura, T., J. Phys. Chem. 84, 365 (1980). 15. Mukerjee, P., and Yang, A., J. Phys. Chem. 80, 1388 (1976). 16. The values of the micelles radii reported in Refs. (4) (quasielastic lightscattefing) and (5, 10) (pulsed gradient Fourier transform NMR) are also ap- proximate, for the reasons given in the text. 17. Atik, S., Nam, M., and Singer, L., Chem. Phys. Lett. 67, 75 (1979). 18. Llanos, P., and Zana, R., J. Colloid Interface Sci. 84, 100 (1981). 19. Lianos, P., Lang, J., Sturm, J., and Zana, R., J. Phys. Chem. 88, 819 (1984) and references therein. 20. Malliaris, A., Lang, J., and Zana, R., J. Phys. Chem., in press. 21. Infelta, P., Gr~tzel, M., and Thomas, J., £ Phys. Chem. 78, 190 (1974). 22. Tachiya, M., Chem. Phys. Lett. 33, 289 (1975). 23. Van der Auweraer, M., Dederen, J., Gdad6, E., and De Schryver, F., J. Phys. Chem. 74, 1140 (1981). 24. Hoyer, H., and Doerr, I., J. Phys. Chem. 68, 3494 (1964). Journal of Colloid and Interface Science, VoL 110, No. 1, March 1986