Langmuir 1996,11, 1855-1857
1855
Cross-Linked Micelles F. M. Menger” and Alexey V . Eliseev Department of Chemistry, Emory University, Atlanta, Georgia 30322 Received February 14, 1995. In Final Form: April 13, 1995@ Micelles experience a dramatic increase in size when mixed with a few molecules per micelle of an oppositely charged gemini surfactant. This size increase was observed via spin-spin relaxation times, dynamic light scattering, and light microscopy. Controlling micelle size with cross-linking agents could have useful practical applications. We recently reported the synthesis and characterization of a new class of self-assembling molecules called “gemini surfactants”.l These substances possess, in sequence, a long hydrocarbon chain, an ionic group, a rigid spacer, a second ionic group, and another hydrocarbon tail. Compounds 1 and 2 provide two examples. When dissolved
A
B
Figure 1. Two possible binding modes for an anionic gemini surfactant associated with a cationic micelle. The dark rectangle of the gemini represents its rigid stilbene spacer.
CH3 2
in water, gemini surfactants display a property unique to surfactant chemistry: Critical micelle concentrations are higher for long-chained gemini surfactants (16-20 carbons) than for shorter-chained analogs (12 carbons). The long-chained geimini surfactants are believed to engage in self-coiling and submicellar aggregation when first exposed to water. Self-assembly into micelles then proceeds as a second step over the course of many hours or days. Although gemini surfactants are obviously interesting in their own right,2they seemed to hold even more promise when admixed with conventional surfactants. Thus, anionic gemini 1 should bind tightly, via electrostatic and hydrophobic forces, to cationic micelles of CTAB. Figure
1shows two possible modes ofbinding. The question arose as to how a few mole percent of gemini 1 admixed with CTAB would affect the properties of the CTAB micelles. In pursuing a n answer to this question, we found, as demonstrated below, that low levels of gemini induce a remarkable micellar growth. The article ends with speculations as to the source of this effect. ‘H NMR relaxation times (TI and Tz) were obtained on the CTAB N-methyl protons using a GN-500 500 MHz spectrometer. CTAB solutions (5.0 mM in 0.05 M NaBr M in DzO, pD = 8.50) contained 1.9 x to 5.6 x gemini 1. Inversion-recoveryand Hahn spin-echomethods
* Abstract published in Advance A C S Abstracts, May 15,1995.
(1)Menger, F. M.; Littau, C. A. J.Am. Chem. SOC.1993,115,10083. See this paper for references to compounds related togemini surfactants. (2) Karaborni, S.;Esselink, K.; Hilbers, P. A. J.;Smit, B., Karthtiuser, J.; van Os, N. M.; Zana, R. Science 1994,266,254.
0743-746319512411-1855$09.00/0
were used to measure T1 and Tz,re~pectively.~ A series of 10-12 delay times (1ms to 10 s for TIand 0.5 ms to 0.5 s for Tz) were applied, and the results were processed by standard spectrometer sofiware. Light scattering studies were performed on a NICOMP submicrometer particle sizer, Model 370, and processed by NICOMP software. Data sets were accumulated for 2-4 h from samples thermostated at 23 “C. Since we were focusing on gemini-induced micelle growth (as opposed to gemini-induced micelle formation), all experiments used CTAB at concentrations considerably M. above its critical micelle concentration4 of 8 x Longitudinal relaxation times ( T I were ) almost unaffected by increasing concentrations of gemini 1. Thus, adding 3.3 x M of 1 to the 5.0 mM solution of CTAB changed the TIof the CTAB N-methyl protons from 482 f 2 to 475 f 9 ms. The constant TIfor the gemini-doped micelles shows that motional freedom within the aggregates (crosslinked or otherwise) has not been p e r t ~ r b e d . ~ Transverse relaxation times (Tz) of the N-methyl protons, on the other hand, decreased dramatically upon addition of small amounts of 1 (Figure 2). For example, M CTAB M gemini 1 to 5.0 x adding 5.6 x diminished Tz from 41 to 1.6 ms. This gemini concentration corresponds to 11 gemini molecules per micelle (assuming a CTAB aggregation number of Even one gemini molecule per micelle has a substantial effect: Tz is decreased 2-fold. Lowered Tz values indicate an impaired tumbling of enlarged self-assembled system^.^ Other CTAB protons showed a similar behavior, but their relaxation times are less reliable owing to signal overlap and broad line widths. Control experiments with a simple anionic surfactant, dodecyl phosphate, showed much smaller effects on Tz. As seen in Figure 3,the TZus concentration plot levels off (3)Sanders, J. K.M.; Hunter, B. K. Modern NMR Spectroscopy; Oxford University Press: Oxford, 1987;pp 61-76. (4)Bany,B.W.; Morrison, J. C.; Russell, G. F. J. J.Colloid Interface Sci. 1970,33, 554. ( 5 ) Levy, G.C. Acc. Chem. Res. 1973,6,161. (6)Roelants, E.;De Schryver, F. C. Langmuir 1987,3, 209. (7)Dwek, R. A. Nuclear Magnetic Resonance in Biochemistry; Clarendon Press: Oxford, 1973;pp 25-27.
0 1995 American Chemical Society
1856 Langmuir, Vol. 11,No. 6,1995
Letters Table 1. Size of CTAB Aggregates with Increasing Concentration of Gemini Surfactant la
4
[lMCTABI
40
0
30
0.016 0.094 0.19
5500
method DLSb DLS light microscopy light microscopy
103- 104 a [CTAB] = 5 mM in 0.05 M NaBr. Dynamic light scattering.
0
E
2
diameter, nm 2-3 4-6
20
e
M dodecyl were unaffected by the addition of 7.4 x phosphate to the 5 mM CTAB. Addition of 7.7 x gemini 1, however, shifted the peaks to 4-6 nm (83%) and 30-35 nm (17%),respectively(Table 1). This doubling of the apparent hydrodynamic diameter is caused by an amount of gemini corresponding to an average of 1.5 molecules per 100 molecules of CTAB. At gemini concentrations of 5-7 mol %, the light scattering data became unstable owing, we believe, to formation oflarge suspended particles that dominated the scattering. Inspection of the suspensions under the light microscope revealed phaseseparated, coacervate-like droplets, the largest of which had diameters of lo4 nm (Table 1). Note that the growth in the micelles far exceeds the total volume of the gemini surfactants. Growth by simple mixed-micelle formation seems, therefore, eliminated as a possibility. Since the gemini cause no obvious viscosity increase, growth into long fibrous aggregates also seems unlikely. Addition of 5 mol % gemini 1 to CTAB solutions of various concentrations produced virtually no change in the critical micelle concentration as determined by surface tension. This is consistent with rather normal micellization followed by gemini-induced growth as opposed to massive molecular reorganization. In summary, we have shown that tiny amounts of gemini surfactant (even one or two molecule per micelle) can promote a substantial growth in micelles of opposite charge. The question arises, of course, as to the source of this effect. Several possibilities come to mind: (a) The micelles might be “cross-linked” as s h o e in Figure 3. Note that this is a highly schematic picture; intermicellar repulsions at the gemini binding-sites could easily be reduced by distortion of the idealized spheres into more elongated structures. Only minor amounts of energy are required to deform spherical micelles into prolate or oblate shape^.^ Actually, the idea of micelle clustering is not new. Blanamer et aZ.1° have provided strong evidence via DSC measurements that micelles of 0.02 M CTAB and three other surfactants exist in the clustered state. We quote from their work: “The consistency for four Surfactants is reassuring and provides good evidence for assemblies of a small number of micelles in these so1utions.”l0 Although an “assemblyof micelles” is not a new concept, this seems to be the first time that micelle clustering has been proposed to arise via a simple additive.” The term “cross-linked”was coined to describe Figure 3, but the model should not be confused with the permanent, covalent cross-linking found in polymeric systems. Micelles, being dynamic entities, would continually cluster, break apart, and recluster on the microsecond time scale.
e 8
lo 0 0
20
40
60
[additlv.],
0
loo
120
rtP u
Figure2. Eff‘ect of dodecyl phosphate (0)and gemini surfactant
1 (a)upon the spin-spin relaxation times of the CTABN-methyl protons (2’2 in ms). The 2’2 values were determined by spinM CTAB (D20,0.05M NaBr, echo experiments using 5.0 x pH 8.5).
Figure 3. Cationic micelles cross-linked by a n anionic gemini surfactant. The drawing is highly stylized and ignores the deformaties from the spherical shape that would arise from intermicellar repulsions.
at 15-20 ms. Thus, 1.2 x M dodecyl phosphate gives a 2’2 of 18 ms which is 11 times the 2’2 given by gemini 1 at half the concentration! A second control experiment with p-dodecylphenylsulfonate (DPS) showed that this monoanionic surfactant can, through a poorly understood mechanism, also enlarge CTAB micelles. But it does so much less efficiently than the gemini. Thus, 1.2 x M DPS and 5.6 x M gemini 1 lower 2’2 to 5.5 and 1.6 ms, respectively. A single NMR experiment with ca. 8 mol % cationic gemini 2, used in plce of 1, increased 2’2 of the CTAB N-methyl signal from 41 ms to about 83 ms. This may be related to formation of mixed micelles of smaller size. Dynamic light scattering of pure CTAB solutions under the conditions specified in Table 1 gave two major peaks with diameters of 2.4 nm (84% by volume) and 13 nm (16% by volume).8 These numbers lie at the detection limits of our laser, and they are primarily for comparison purposes only. The position and intensity of the peaks ( 8 )The bimodal distribution, typical of dynamic light scattering output, reflects the polydispersity of the system.
(9)Walderhaugh, H.;Sliderman, 0.;Stilbs,P. J.Phys. Chem. 1984, 88, 1655. (10)Blandamer, M.J.; Briggs, B.; Burgess, J.; Butt, M. D.; Brown, H. R.;Cullis, P. M. J. Colloid Interface Sci. 1992, 150, 285. (11)Cross-linking of micelles by polymers is a well-documented phenomenon: Sarrazin-Cartalas,A.; Iliopoulos,I.;Audebert,R.;Olsson, U. Langmuir 1994,10, 1421. See Figure 2.
Letters (b)It is possible that the gemini bind in an intramicellar fashion as shown in Figure 1B. One or more gemini surfactants could create a “hydrophobicpatch” comprised of stilbene spacers and various CT& chain segments that happen to reside at the micelle surface. Association of the micelles could then take place via hydrophobic attraction between the “fatty lesions”,electrostatic repulsion notwithstanding. In principle, even a properly designed nongemini might operate by this mechanism. (c) Finally, one cannot discount a model in which a gemini induces fusion of two micelles, leaving the gemini buried in the middle of the new aggregate. Accordingly, a single gemini molecule would have a profound effect on the overall micellar morphology and curvature. Although such a molecular reorganization might seem at odds with the critical micelle concentration results, it remains a possibility.
Langmuir, Vol. 11, No. 6, 1995 1857 The three models for micellar growth are all variations on a theme, differing mainly in the contact distance between neighboring micelles. Further experiments to differentiate the models are desirable (especially so since the area of micellar structure has seen little movement followingthe abandonment of the Hartley “wheel-spoke”” model many years ago12). Regardless of mechanism, the possibility now exists of controllingmicellar size, and perhaps micellar properties, using relatively small amounts of additive. Given the ubiquitous applications of surfactants in industry, this could be a useful development. Acknowledgment. This work was supported by the National Institutes of Health. LA950109Y ~
(12)Menger, F. M.Acc. Chem. Res. 1979, 12, 111
