Langmuir 2008, 24, 10771-10775
10771
Mechanism of Surfactant Micelle Formation Xiaohong Cui,†,‡ Shizhen Mao,*,† Maili Liu,*,† Hanzhen Yuan,† and Youru Du† State Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics, Wuhan Institute of Physics and Mathematics, Chinese Academy of Sciences, Wuhan 430071, P. R. China, Graduate School, Chinese Academy of Science, Beijing 100029, P. R. China ReceiVed June 3, 2008. ReVised Manuscript ReceiVed June 28, 2008 The mechanism of micelle formation of surfactants sodium dodecyl sulfate (SDS), n-hexyldecyltrimethylammonium bromide (CTAB) and Triton X-100 (TX-100) in heavy water solutions was studied by 1H NMR (chemical shift and line shape) and NMR self-diffusion experiments.1H NMR and self-diffusion experiments of these three surfactants show that their chemical shifts (δ) begin to change and resonance peaks begins to broaden with the increase in concentration significantly below their critical micelle concentrations (cmc’s). At the same time, self-diffusion coefficients (D) of the surfactant molecules decrease simultaneously as their concentrations increase. These indicate that when the concentrations are near and lower than their cmc’s, there are oligomers (premicelles) formed in these three surfactant systems. Carefully examining the dependence of chemical shift and self-diffusion coefficient on concentration in the region just slightly above their cmc’s, one finds that the pseudophase transition model is not applicable to the variation of physical properties (chemical shift and self-diffusion coefficient) with concentration of these systems. This indicates that premicelles still exist in this concentration region along with the formation of micelles. The curved dependence of chemical shift and self-diffusion coefficient on the increase in concentration suggests that the premicelles grow as the concentration increases until a definite value when the size of the premicelle reaches that of the micelle, i.e., the system is likely dominated by the monomers and micelles. Additionally, the approximate values of premicelle coming forth concentration (pmc) and cmc were obtained by again fitting chemical shifts to reciprocals of concentrations at a different perspective, and are in good accordant with experimental results and literature values and prove the former conclusion.
Introduction Surfactant molecules, which are made up of polar, ionic, or nonionic head groups and extended apolar, organic residues, in aqueous solution self-assemble into micelles at concentrations above the so-called critical micelle concentration (cmc). There is a longstanding controversy in the mechanism of surfactant micelle formation. The mass-action and the pseudophase transition models are the two most common models used in describing the micellization process.1-3 The former considers the micelle formation as a multistep process, and it is able to explain the gradual variation of physical properties of the system in the concentration range near the cmc. At concentrations below the cmc, there is a distribution of monomers and premicelles. The latter assumes that micelles are formed in a single process and predicts an abrupt change of the solution physical properties at the cmc.3 Up to now, numerous experimental and theoretical papers reported the appearance of aggregates for some gemini surfactants at concentrations below the cmc’s, a phenomenon referred to as premicellar aggregation.4-13 Menger et al.4 first investigated the aggregation of the geminis by surface tension, film-balance methods, dynamic light scattering, lH and 23Na NMR, and spectral changes in adsorbed dye. Later, the Rosen group5,6 studied the micellization and premicellar aggregation of some gemini surfactants by surface tension, interfacial tension, and fluorescence methods and gave remarkable evidence of premicelles of the * Corresponding author. E-mail address:
[email protected] (S.Z.M.);
[email protected] (M.L.L.). Tel.:86-27-87197305. Fax: 86-27-87199291. † Wuhan Institute of Physics and Mathematics. ‡ Graduate School. (1) Blandamer, M. J.; Cullis, P. M.; Soldi, L. G.; Engberts, J. B. F. N.; Kacperska, A.; Van Os, N. M.; Subha, M. C. S. AdV. Colloid Interfaces 1995, 58(2-3), 171–209. (2) Wettig, S. D.; Verrall, R. E. J. Colloid Interface Sci. 2001, 235(2), 310– 316. (3) Amato, M. E.; Caponetti, E.; Martino, D. C.; Pedone, L. J. Phys. Chem. B 2003, 107(37), 10048–10056.
studied gemini surfactants. Recently, Hadgiivanova and Diamant14 probed into amphiphilic molecule aggregating using a simple two-state (monomer-aggregate) thermodynamic model, presenting that, at a concentration range below the cmc, a significant fraction of amphiphilic molecules participate in aggregates, and the aggregate size is found to weakly change with concentration below and above the cmc, which provides strong theoretical support for premicellar aggregation for amphiphilic molecules. Lately, an unexpected peak in light scattering as the concentration varies between zero and above the cmc was observed near the cmc of surfactants n-hexyldecyltrimethylammonium bromide (CTAB, cationic surfactant), sodium dodecyl sulfate (SDS, anionic surfactant), and Triton X-100 (TX-100, nonionic surfactant) using static light scattering.15 However, this property appears to be one that has not previously been observed, and they did not explain this phenomenon. As far as we know, it is probably attributed to premicellar aggregation. Nuclear magnetic resonance (NMR) is a powerful method in studying surfactant (4) Menger, F. M.; Littau, C. A. J. Am. Chem. Soc. 1993, 115(22), 10083– 10090. (5) Song, L. D.; Rosen, M. J. Langmuir 1996, 12(5), 1149–1153. (6) Rosen, M. J.; Mathias, J. H.; Davenport, L. Langmuir 1999, 15(21), 7340– 7346. (7) Mathias, J. H.; Rosen, M. J.; Davenport, L. Langmuir 2001, 17(20), 6148– 6154. (8) Mackie, A. D.; Panagiotopoulos, A. Z.; Szleifer, I. Langmuir 1997, 13(19), 5022–5031. (9) Zana, R. AdV. Colloid Interfaces 2002, 97(1-3), 205–253. (10) Wettig, S. D.; Wang, C. Z.; Verrall, R. E.; Foldvari, M. Phys. Chem. Chem. Phys. 2007, 9(7), 871–877. (11) Li, X. F.; Wettig, S. D.; Wang, C. Z.; Foldvari, M.; Verrall, R. E. Phys. Chem. Chem. Phys. 2005, 7(17), 3172–3178. (12) Waengnerud, P.; Joensson, B. Langmuir 1994, 10(10), 3542–3549. (13) Jiang, Y.; Chen, H.; Cui, X. H.; Mao, S. Z.; Liu, M. L.; Luo, P. Y.; Du, Y. R. Langmuir 2008, 24(7), 3118–3121. (14) Hadgiivanova, R.; Diamant, H. J. Phys. Chem. B 2007, 111(30), 8854– 8859. (15) Sorci, G. A.; Walker, T. D. Langmuir 2005, 21(3), 803–806.
10.1021/la801705y CCC: $40.75 2008 American Chemical Society Published on Web 08/26/2008
10772 Langmuir, Vol. 24, No. 19, 2008
Cui et al.
Figure 1. Chemical structures and proton numbering of the three surfactants studied.
aggregating. Changes in chemical shifts, line widths, and line shapes of 1H NMR spectrum due to alteration in the chemical environments of surfactant molecules provide direct and strong evidence of premicellar aggregation.16 Self-diffusion coefficients quantify the size of surfactant molecules in aqueous solutions, which reflects the state in which the molecules are. Therefore, the micellization of these three types of surfactants were investigated by 1H NMR and NMR self-diffusion experiments.
Experimental Section Materials. The SDS (MW ) 288.38) was purchased from Alfa Aesar with a purity of 99%. TX-100 (MW ) 646.86) was purchased from Nacalai Tesque to ensure the highest purity. The CTAB (MW ) 364.45) and D2O were purchased from Acros with a purity of 99% and a deuteration of 99.8%. The above-mentioned reagents were used as received, without any further purification. D2O was used as solvent instead of water in order to weaken the water signal. Experiments. All NMR experiments were performed at 25 °C on a Bruker AVANCE spectrometer with a proton frequency of 500.13 MHz, except 1H NMR experiments of TX-100 which were performed on Varian INOVA-600 at a proton frequency of 599.82 MHz. Me3Si-CD2CD2-CO2Na (TSP) was used as the external reference. For assuring complete recovery of magnetization vector, a small pulse flip-angle 30° was used rather than 90° in the conventional single pulse sequence. The bipolar gradient longitudinal eddy-current delay (LED-BPP) sequence17 was used to measure the self-diffusion coefficient, D.
Figure 2. Selected regions of 1H NMR spectra of TX-100 at various concentrations in D2O at 25 °C, cmc ) 0.30 mM.
Results and Discussion Chemical structures and proton numbering of the three surfactants studied are shown in Figure 1. The selected regions of 1H NMR spectra of TX-100 at various concentrations in D2O at 25 °C are shown in Figure 2. It is evident that the 1H NMR spectra at concentrations of 0.19 and 0.22 mM (marked in black) are identical. It shows that the surfactant molecules in solution remain in the monomeric state at concentrations below its cmc. However, at the concentration range of 0.23-0.28 mM (marked in green), which is still below and almost near the cmc (0.30 mM18) for TX-100, the chemical shifts (δ) of resonance peaks for the phenyl protons move to a higher field as the concentration increases, and the split double resonance peaks begin to merge together, changing from sharp and narrow to smooth and wide. Shifting to a higher field for δ implies that the environment of the monomer molecules is varied, because of the phenyl ring current effect resulting from the association of monomers as the phenyl rings come close together. The merging and broadening of resonance peaks also result from aggregating of monomers and exchange of monomers in between premicelles and in bulk solution. Thus, 1H NMR spectra give strong evidence about the formation of (16) Gillitt, N. D.; Savelli, G.; Bunton, C. A. Langmuir 2006, 22(13), 5570– 5571. (17) Wu, D. H.; Chen, A. D.; Johnson, C. S. J. Magn. Reson., Ser. A 1995, 115(2), 260–264. (18) Fendler, J. H. Membrane Mimetic Chemistry. Characterization and Application of Micelles, Microemulsions, Monolayers, Bilayers, Vesicles, HostGuest Systems and Polyions; John Wiley & Sons; New York, 1982; p 69.
Figure 3. Selected regions of 1H NMR spectra of CTAB at various concentrations in D2O at 25 °C, cmc ) 0.92 mM. Table 1. Chemical Shifts δobsd (ppm) for Various Protons and Self-Diffusion Coefficients Dobsd (10-10 m2 · s-1) of the Aqueous Solutions of SDS at Various Concentrations in D2O at 25 °C, cmc ) 8.40 mM δobsd C (mM) 3.44 4.37 5.21 5.96 6.72 7.39 8.40 10.16 11.76 16.88 19.66 45.61
H1
H5
Dobsd
0.9220 0.9219 0.9221 0.9225 0.9227 0.9251 0.9273 0.9349 0.9439 0.9461 0.9497 0.9592
4.1231 4.1228 4.1230 4.1232 4.1225 4.1218 4.1201 4.1148 4.1086 4.1049 4.1024 4.0977
5.042 5.035 5.026 5.015 4.779 4.556 3.743 3.037 2.450 2.083 1.045
premicelles for the nonionic surfactant, TX-100, at concentrations below its cmc. 1H NMR spectra of the cationic surfactant, CTAB, at various concentrations in D2O at 25 °C show similar phenomena with
Mechanism of Surfactant Micelle Formation
Langmuir, Vol. 24, No. 19, 2008 10773
Figure 4. Variations of δobsd as a function of reciprocals of concentrations in D2O at 25 °C: (a) SDS (H1), (b) CTAB (H6). The points of intersection of two lines indicate cmc’s.
those of TX-100 (Figure 3). The 1H NMR spectra at concentrations of 0.49 and 0.65 mM (marked in black) are identical. When its concentration reaches 0.74 mM (marked in green), the chemical shifts of resonance peaks of both N-methyl protons (H6) and terminal methyl protons (H1) start to shift to a lower field, and the half-line width of H6 becomes wider. These indicate that the surfactant molecules with mutual static-effecting begin to be close to each other, at concentrations below the cmc, i.e., they aggregate into oligomers (premicelles, the aggregate number of which is less than that of micelles). The chemical shifts move further to the high-frequency side as the concentration increases (spectrum at 0.83 mM). When the concentration reaches its cmc, 0.92 mM18 (marked in red), both the chemical shift and the line shape change significantly with the increase in concentration, which implies micelle formation. The variation in chemical shift (δobsd) and self-diffusion coefficients (Dobsd) of the anionic surfactant, SDS, shown in Table 1, gives the same conclusion. Consequently, the premicellar aggregation for these three types of surfactants well explain the unexpected peak in light scattering at concentrations below and near their cmc’s.15 For surfactants of fast exchange between monomers and micelles in bulk solution, the observed chemical shift (δobsd) of the resonance peak can be expressed as the weighted mean of chemical shifts of the micelles and monomers at concentrations above their cmc’s according to the pseudophase transition model1-3 by the following equation:
δobsd ) (Cmon/CT)δmon + (Cmic/CT)δmic
(1)
where δmon and δmic represent the chemical shifts related to the free monomers and to the monomers in micelles, respectively; Cmon and Cmic are the free surfactant concentration and the micellized surfactant concentration, respectively; and CT ) Cmon + Cmic is the total surfactant concentration. Below the cmc, the pseudophase transition model2 predicts that δmon remains constant and equal to δobsd. Above the cmc, it is supposed that the free monomer concentration remains constant as the cmc, so eq 1 can be rewritten as
δobsd ) (cmc/CT)(δmon - δmic) + δmic
(2)
According to eq 2, δobsd at concentrations above and below the cmc should form two straight lines, and the point of intersection of these two lines is the cmc.
Figure 5. Variations of self-diffusion coefficients (Dobsd) of SDS molecules as a function of reciprocals of concentrations in D2O at 25 °C. The points of intersection of two lines indicate the cmc.
The observed proton chemical shifts (δobsd) of the terminal methyl group (SDS) and of the N-methyl group (CTAB) as a function of reciprocals of concentrations are shown in Figure 4a,b, respectively. We can really draw the two straight lines for the two respective surfactants, and the two points of intersection, the cmc’s, agree well with the respective literature values. However, not all δobsd of resonance peaks fall on these two lines. At concentration ranges slightly below or above the cmc, δobsd values can not follow the formula and are higher than the fitting lines. We have discussed in the previous section that those premicelles with higher chemical shift values are formed at concentrations below the cmc, resulting in a higher weighted mean chemical shift. But the deviation of δobsd dependence on reciprocals of concentrations at concentrations just above the cmc from eq 2 should receive attention. This may be attributed to the existence of premicelles in the bulk solution at concentrations just above the cmc, the chemical shift of which is larger than that of the monomer, causing a higher weighted mean chemical shift. That is to say, in this concentration region, there are monomers, oligomers, and micelles coexisting in the solutions.
10774 Langmuir, Vol. 24, No. 19, 2008
Cui et al.
Figure 6. The fitting graphs of δobsd to reciprocals of concentrations for SDS (a) and CTAB (b) according to eqs 4 and 6.
Carefully examining the dependence of chemical shifts on concentration in the region from just slightly below to just slightly above their cmc’s, one finds that the experimental points form a curve rather than a straight line. It implies that the size of premicelles may grow as the concentration increases. The molecular size can be reflected by its self-diffusion coefficients (D). Just as in the case of chemical shift, dependence of D on concentration can also be described by the pseudophase transition model.1-3 Variations in self-diffusion coefficients (Dobsd) of surfactant molecules for SDS as a function of reciprocals of concentrations in D2O at 25 °C are shown in Figure 5. At concentration ranges from 5.96 to 7.14 mM, which are below the cmc (cmc ) 8.40 mM15), the self-diffusion coefficients (D) of H2O remain constant, which shows that there are almost no changes in the viscosity of solutions. However, Dobsd of the SDS molecules decrease with concentration increasing from 5.96 mM. This decrease could not be attributed to any variation of the system viscosity. It is direct evidence of the aggregation of the surfactant, i.e., the formation of premicelles at concentrations below the cmc, and they are growing as the concentration increases. At concentrations a little above the cmc, the variation of Dobsd does not obey the pseudophase transition model either. It strongly supports the results obtained from the chemical shift variations. Consequently, the micellization is a multistep process, i.e., the surfactant monomers in the solution start to associate into small aggregates at a definite concentration below the cmc, they grow to form larger aggregates as the concentration increases, micelles form at the cmc, where premicelles still exist, and finally, at a definite concentration, a little above the cmc, when the premicelles grow into the size of the micelles, the system is likely dominated by the monomers and micelles in the bulk solution. Let us think about it from a different perspective. When premicelles generate at concentrations below the cmc, the observed chemical shift (δobsd) of the resonance peak can be expressed as the weighted mean of chemical shifts of monomers and premicelles by another equation, eq 3, which is similar to eq 1, where δmon and δpre represent the chemical shifts related to the free monomers and the monomers in premicelles, respectively; Cmon and Cpre are the free surfactant concentration and the premicellized surfactant concentration, respectively; and CT ) Cmon + Cpre is the total surfactant concentration. Likewise,
Table 2. The Fitting Results (Intercept and Slope) of δobsd to Reciprocals of Concentrations for SDS and CTAB as Described in Figure 6 SDS_H1 parameter a
intercept slopea interceptb slopeb a
SDS_H5
value
error
value
0.9254 -0.0137 0.9661 -0.3122
0.0029 0.0146 0.0019 0.0240
4.1217 0.0054 4.0914 0.2308
error
CTAB_H6 value
error
0.0018 3.2673 0.0020 0.0090 -0.0863 0.0033 0.0012 3.1687 0.0044 0.0148 -0.0039 0.0024
Fitting values obtained from eq 4. b Fitting values obtained from eq 6.
it is supposed that δmon remains constant and equals δobsd, which remains constant at lower concentrations, and Cmon, the free monomer concentration, remains constant as the premicelle coming forth concentration, named the pmc, so eq 3 can be rewritten as eq 4.
δobsd ) (Cmon ⁄ CT)δmon + (Cpre ⁄ CT)δpre
(3)
δobsd ) (pmc ⁄ CT)(δmon - δpre) + δpre
(4)
In the same way, at concentrations slightly above the cmc, the observed chemical shifts (δobsd) are contributed by those of monomers, premicelles, and micelles, as described by eq 5, where all parameters are defined as above and will not be explained repeatedly; Cmon and Cpre are thought of as constants and equal the pmc and cmc - pmc, respectively; and CT ) Cmon + Cpre + Cmic. Cmic is equal to CT - cmc; therefore, we have eq 5 and eq 6.
δobsd ) (Cmon ⁄ CT)δmon + (Cpre ⁄ CT)δpre + (Cmic ⁄ CT)δmic (5) δobsd ) (pmc ⁄ CT)(δmon - δpre) + (cmc ⁄ CT)(δpre - δmic) + δmic (6) Because the proportion of premicelles in solution is very small, to simplify the fitting process, the approximate pmc and cmc can be gained by fitting all of the observed chemical shifts (δobsd) at concentrations below and above the cmc to reciprocals of the concentrations according to eqs 4 and 6, respectively. The fitting graphs and results are listed in Figure 6 and Table 2, respectively. According to eqs 3 and 5, and the fitting results listed in Table 2, the values of pmc and cmc obtained for SDS and CTAB are shown in Table 3. From Table 3, it can be found that the values
Mechanism of Surfactant Micelle Formation
Langmuir, Vol. 24, No. 19, 2008 10775
Table 3. The Values of pmc and cmc Obtained for SDS and CTAB According to Eqs 3 and 5 and the Fitting Results in Table 2 pmc (mM) cmc (mM)
SDS_H1
SDS_H5
CTAB_H6
4.0 7.3
4.2 7.4
0.45 0.84
of cmc’s are in the same order of magnitude with literature values. However, it is also obvious that the cmc’s of both SDS (7.4 mM) and CTAB (0.84 mM) are about 12% and 9% lower than the reported values (8.4 mM15 for SDS and 0.92 mM18 for CTAB), respectively. The cmc value thus obtained (0.84 mM) for CTAB agrees well with the value (0.82 mM) determined by using the light scattering approach.15 The above-mentioned differences may be due to the methodology and experimental error used to define the cmc.19 As for pmc’s, the values derived from the chemical shift data are 4.1 mM and 0.45 mM for SDS and CTAB, respectively. Additionally, other useful information is available from Table 3, viz., the relative amounts of surfactant molecules in the three states (monomer, premicelle, and micelle) at several concentrations (at, below, and above the cmc) listed in Table 4. It tells not only the relative proportions of surfactant molecules in each state at a certain concentration, but also variations of the relative amounts as the concentration increases. We have to indicate that the pmc derived from chemical shift data may contain larger error than that of cmc because of the small difference between chemical shifts of monomers and premicelles. Nevertheless, this provides useful information of the concentration at which the surfactant monomers in the solution start to associate into premicelles below the cmc. For a similar reason, at concentrations above the cmc, the chemical shift and self-diffusion coefficient (19) Zana, R.; Benrraou, M.; Rueff, R. Langmuir 1991, 7(6), 1072–1075.
Table 4. The Relative Amounts of Surfactant Molecules in the Three States (Monomer, Premicelle, and Micelle) at Several Concentrations SDS C (mM) 4.37 7.39 8.40 10.16 45.61
proportion (%)
CTAB
proportion (%)
Pmon
Ppre
Pmic
C (mM)
Pmon
Ppre
Pmic
94 55 49 40 9
6 45 39 33 7
12 27 84
0.49 0.83 0.92 1.20 7.39
92 54 49 38 6
8 46 42 32 5
9 30 89
of the micelles dominate these two NMR-observed parameters, and the effect of premicelles could be ignored. Therefore, for data analysis, we may only consider the micelles and monomers in the bulk solution.
Conclusions The micellization of three types surfactants in aqueous solution has been investigated by 1H NMR spectra and NMR self-diffusion experiments. Changes in chemical shifts, line shape, and line width in 1H NMR spectra and variations in self-diffusion coefficients at concentrations near and below the cmc show that premicelles start to generate and grow into larger ones as the concentration increases. When the concentration reaches the cmc, micelles come into being, but the premicelles still exist and grow until their size is comparable with that of the micelles. In conclusion, the micelle formation is a multistep and gradual process. The approximate values of pmc and cmc obtained by fitting chemical shifts to reciprocals of concentrations also support this conclusion. Acknowledgment. Financial support by the National Science Foundation of China (20610104, 20635040) is gratefully acknowledged. LA801705Y
