Fluorescence Delineation of the Surfactant Microstructures in the

and anionic surfactants exhibit a rich array of self-assembled microstructures ... Kerry K. Karukstis , Whitney C. Duim , Gerald R. Van Hecke , an...
0 downloads 0 Views 297KB Size
Download PDF
64

Langmuir 2004, 20, 64-72

Fluorescence Delineation of the Surfactant Microstructures in the CTAB-SOS-H2O Catanionic System Kerry K. Karukstis,* Shelley A. McCormack, Tyrel M. McQueen, and Kimberly F. Goto Department of Chemistry, Harvey Mudd College, Claremont, California 91711 Received July 1, 2003. In Final Form: October 14, 2003 A key feature of amphiphilic molecules is their ability to undergo self-assembly, a process in which a complex hierarchical structure is established without external intervention. Ternary systems consisting of aqueous mixtures of cationic and anionic surfactants exhibit a rich array of self-assembled microstructures such as spherical and rodlike micelles, unilamellar and multilamellar vesicles, planar bilayers, and bicontinuous structures. In general, multiple complementary techniques are required to explore the phase behavior and morphology of aqueous systems of oppositely charged surfactants. As a novel and effective alternative approach, we use fluorescence spectroscopic measurements to examine the microstructures of aqueous cationic/anionic surfactant systems in the dilute surfactant region. In particular, we demonstrate that the polarity-sensitive fluorophore prodan can be used to demarcate the surfactant microstructures of the ternary system of cetyltrimethylammonium bromide, sodium octyl sulfate, and water. As the fluorescence signature of this probe is dependent on the nature of the surfactant aggregates present, our method is a promising new approach to effectively map complex surfactant phase diagrams.

Introduction A variety of surfactant aggregates are present in aqueous solutions of binary surfactant mixtures. One such interesting system is that of two oppositely charged surfactants in aqueous solution. A strong synergism arises from electrostatic interactions between the amphiphiles and results in a complexity of microstructures far greater than exhibited by the individual surfactants. Factors such as relative alkyl chain lengths, number of alkyl chains per surfactant, total surfactant concentration, surfactant mixing ratio, and temperature dictate the rich array of aggregates formed.1-13 Figure 1 illustrates the multiplicity of aggregates formed at room temperature by the ternary system of water and two surfactants with highly asymmetric alkyl chain lengths, cationic cetyltrimethylammonium bromide (CTAB, C16) and anionic sodium octyl sulfate (SOS, C8).4 The dilute surfactant region illustrated * To whom correspondence should be addressed. E-mail: [email protected]. Telephone: (909) 607-3225. Fax: (909) 607-7577. (1) Kaler, E. W.; Herrington, K. L.; Murthy, A. K.; Zasadzinski, J. A. N. J. Phys. Chem. 1992, 96, 6698-6707. (2) Brasher, L. L.; Herrington, K. L.; Kaler, E. Langmuir 1995, 11, 4267-4277. (3) Filipovic-Vincekovic, N.; Bujan, M.; Dragcevic, D.; Nekic, N. Colloid Polym. Sci. 1995, 273, 182-188. (4) Yatcilla, M. T.; Herrington, K. L.; Brasher, L. L.; Kaler, E. W. J. Phys. Chem. 1996, 100, 5874-5879. (5) Caria, A.; Khan, A. Langmuir 1996, 12, 6282-6290. (6) Regev, O.; Khan, A. J. Colloid Interface Sci. 1996, 182, 95-109. (7) O’Connor, A. J.; Hatton, T. A.; Bose, A. Langmuir 1997, 13, 69316940. (8) Talhout, R.; Engberts, J. B. F. N. Langmuir 1997, 13, 50015006. (9) Meagher, R. J.; Hatton, T. A.; Bose, A. Langmuir 1998, 14, 40814087. (10) Marques, E. F.; Regev, O.; Khan, A.; Miguel, M. d. G.; Lindman, B. J. Phys. Chem. B 1999, 103, 8353-8363. (11) Villeneuve, M.; Kaneshina, S.; Imae, T.; Aratono, M. Langmuir 1999, 15, 2029-2036. (12) Xiao, J. X.; Sivars, U.; Tjerneld, F. J. Chromatogr., B 2000, 743, 327-338. (13) Minardi, R. M.; Schulz, P. C.; Vuano, B. Colloids Surf. A 2002, 197, 167-172.

has a maximum total surfactant composition of 5 wt %. Observed microstructures include spherical and rodlike mixed micelles, unilamellar vesicles, planar lamellar structures (including liquid crystalline phases), and a crystalline precipitate of equimolar surfactant concentrations. The appearance of stable unilamellar vesicles in mixtures of oppositely charged surfactants has received considerable attention.1,2,4-38 Hexagonal,1,5,13,34,39-41 mi(14) Kaler, E. W.; Murthy, A. K.; Rodriguez, B. E.; Zasadzinski, J. A. N. Science 1989, 245, 1371-1374. (15) Fukuda, H.; Kawata, K.; Okuda, H.; Regen, S. L. J. Am. Chem. Soc. 1990, 112, 1635-1637. (16) Ambuhl, M.; Bangerter, F.; Luisi, P. L.; Skrabal, P.; Watzke, H. J. Langmuir 1993, 9, 36-38. (17) Marques, E.; Khan, A.; Miguel, M. D.; Lindman, B. J. Phys. Chem. 1993, 97, 4729-4736. (18) Herrington, K. L.; Kaler, E. W.; Miller, D. D.; Chiruvolu, S.; Zasadzinski, J. A. J. Phys. Chem. 1993, 97, 13792-13802. (19) Hoffmann, H.; Thunig, C.; Schmiedel, P.; Munkert, U. Langmuir 1994, 10, 3972-3981. (20) Andelman, D.; Kozlov, M. M.; Helfrich, W. Europhys. Lett. 1994, 25, 231-236. (21) Kondo, Y.; Uchiyama, H.; Yoshino, N.; Nishiyama, K.; Abe, M. Langmuir 1995, 11, 2380-2385. (22) Huang, J. B.; Zhao, G. X. Colloid Polym. Sci. 1995, 273, 156164. (23) Zhao, G. X.; Yu, W. L. J. Colloid Interface Sci. 1995, 173, 159164. (24) Oberdisse, J.; Couve, C.; Appell, J.; Berret, J. F.; Ligoure, C.; Porte, G. Langmuir 1996, 12, 1212-1218. (25) Brasher, L. L.; Kaler, E. W. Langmuir 1996, 12, 6270-6276. (26) Walker, S. A.; Zasadzinski, J. A. Langmuir 1997, 13, 50765081. (27) Soderman, O.; Herrington, K. L.; Kaler, E. W.; Miller, D. D. Langmuir 1997, 13, 5531-5538. (28) Laughlin, R. G. Colloids Surf., A 1997, 128, 27-38. (29) Iampietro, D. J.; Brasher, L. L.; Kaler, E. W.; Stradner, A.; Glatter, O. J. Phys. Chem. B 1998, 102, 3105-3113. (30) Filipovic-Vincekovic, N.; Bujan, M.; Smit, I.; Tusek-Bozic, L.; Stefanic, I. J. Colloid Interface Sci. 1998, 201, 59-70. (31) Zhao, G. X.; Yu, W. L.; Gong, Y. J.; Zhu, B. Y. Chin. Chem. Lett. 1998, 9, 1059-1062. (32) Tomasic, V.; Stefanic, I.; Filipovic-Vincekovic, N. Colloid Polym. Sci. 1999, 277, 153-163. (33) Regev, O.; Marques, E. F.; Khan, A. Langmuir 1999, 15, 642645. (34) Marques, E. F. Langmuir 2000, 16, 4798-4807.

10.1021/la0351764 CCC: $27.50 © 2004 American Chemical Society Published on Web 11/25/2003

Delineation of Surfactant Microstructures

Langmuir, Vol. 20, No. 1, 2004 65

Figure 1. Microstructures formed at room temperature for the ternary system of CTAB-SOS-H2O. R ) rodlike micelles, M ) spherical micelles, VC ) cationic-rich unilamellar vesicles, V ) anionic-rich unilamellar vesicles, L ) planar bilayers, and I ) an unidentified multiaggregate region. On the equimolar line, a 1:1 crystalline precipitate forms. Adapted after ref 4.

cellar cubic,34 and bicontinuous cubic5 phases have also been observed in aqueous “catanionic” surfactant mixtures. The CTAB/SOS/H2O aggregates identified in Figure 1 were characterized via visual observation, tensiometry, quasielastic light scattering, pulsed field gradient NMR spectroscopy, and cryogenic transmission electron microscopy (cryo-TEM) using more than 250 ternary mixtures over several months to ensure thermodynamic equilibrium.4 Surface tension measurements defined micellar regions. Visual observations of increased solution viscosity and viscoelasticity distinguished rodlike and spherical micelles. Single-phase vesicle regions and biphasic micellar-vesicle regions were delineated by a combination of visual observations (scattering, viscosity, and turbidity), quasielastic light scattering measurements, and direct imaging via cryo-TEM. The observation of birefringence identified liquid crystalline lamellar phases. The multiaggregate regions containing vesicles were deduced from the high degree of polydispersity in vesicle size revealed by quasielastic light scattering and the lack of a reproducible equilibrium aggregate size after 6 months. The presence of spherical micelles was deduced, for example, by a combination of (1) solution surface tension (γ) measurements as a function of surfactant concentration (c) that revealed a “break” in the linear dependence of γ versus log c followed by (2) 256-1024 NMR scans per sample to monitor the spin-echo signal decay associated with surfactant self-diffusion. Both NMR surfactant self-diffusion measurements and direct imaging via cryo-TEM were required to delineate the region of unilamellar vesicles and the region of micelle-vesicle coexistence. (35) Caillet, C.; Hebrant, M.; Tondre, C. Langmuir 2000, 16, 90999102. (36) Jung, H. T.; Coldren, B.; Zasadzinski, J. A.; Iampietro, D. J.; Kaler, E. W. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 1353-1357. (37) Hao, J. C.; Hoffmann, H.; Horbaschek, K. Langmuir 2001, 17, 4151-4160. (38) Mao, M.; Huang, J.; Zhu, B.; Yin, H.; Fu, H. Langmuir 2002, 18, 3380-3382. (39) Edlund, H.; Sadaghiani, A.; Khan, A. Langmuir 1997, 13, 49534963. (40) Minardi, R. M.; Schulz, P. C.; Vuano, B. Colloid Polym. Sci. 1998, 276, 589-594. (41) Marques, E. F.; Regev, O.; Edlund, H.; Khan, A. Langmuir 2000, 16, 8255-8262.

Figure 2. Chemical structure of the fluorescence probe prodan.

As a novel and effective alternative approach, we use fluorescence spectroscopic measurements to characterize the supramolecular aggregates formed in aqueous cationic/ anionic surfactant systems in the dilute surfactant region. The powerful spectral sensitivity of a fluorescence probe to its environment enables this approach. The fluorophore prodan (6-propionyl-2-(dimethylamino)-naphthalene, Figure 2) is an ideal probe for examining the phase behavior of aqueous cationic/anionic surfactant systems. Prodan’s solubility in a wide range of solvents enables its distribution into an array of single-phase and multiphasic regions.42-45 Furthermore, the observed fluorescence signal can be simultaneously indicative of multiple aggregates as a consequence of the measurable fluorescence intensity of prodan in a range of solvents and the sensitivity of the wavelength of maximum emission (λmax) to the polarity of prodan’s environment.42-45 For example, the emission λmax ranges from 402 nm in the nonpolar solvent cyclohexane to 522 nm in aqueous solution.42,44-46 The fluorescence signature of prodan in surfactant aggregates is a composite spectrum with contributions from the probe population within various microdomains of an aggregate. Deconvolution of the overall prodan fluorescence emission spectrum into a sum of overlapping Gaussian functions enables an interpretation of the location of prodan within discrete microdomains of different relative polarities. We have used prodan to characterize the noncovalent interactions that influence the partitioning of guest molecules within surfactant aggregates such as aqueous micelles,42,44,45 reverse (42) Karukstis, K. K.; Suljak, S. W.; Waller, P. J.; Whiles, J. A.; Thompson, E. H. Z. J. Phys. Chem. 1996, 100, 11125-11132. (43) Karukstis, K. K.; Frazier, A. A.; Martula, D. S.; Whiles, J. A. J. Phys. Chem. 1996, 100, 11133-11138. (44) Karukstis, K. K.; Frazier, A. A; Loftus, C. T.; Tuan, A. S. J. Phys. Chem. B 1998, 102, 8163-8169. (45) Karukstis, K. K. In Handbook of Surfaces and Interfaces of Materials; Nalwa, H. S., Ed.; Academic Press: San Diego, 2001; Vol. 3, Chapter 12. (46) Weber, G.; Farris, F. J. Biochemistry 1979, 18, 3075-3078.

66

Langmuir, Vol. 20, No. 1, 2004

micelles,43-45 and unilamellar vesicles45,47 and within supramolecular assemblies of polymeric dendrimers and surfactants.48 The polarity and water accessibility of the microregions of surfactant aggregates differ considerably and give rise to distinct emission λmax. The fluorescence signature of prodan in these surfactant aggregates suggests that several intermolecular forces influence probe distribution, including hydrophobic interactions, van der Waals (i.e., dipole-dipole) forces, and electrostatic association between the π electrons of the aromatic ring of prodan and cationic surfactant headgroups. Interestingly, while the Gaussian components observed for various surfactant aggregates have similar wavelength maxima, the relative contributions of the Gaussian curves to the overall fluorescence spectrum can vary dramatically with the nature of the aggregate. The distribution pattern for prodan within aggregate microregions is reflected in the fluorescence spectral shape, suggesting that the overall fluorescence spectrum could be used to delineate the different solution-phase aggregates present in an aqueous system of oppositely charged surfactants. To test this hypothesis, we have analyzed the prodan fluorescence spectrum for a range of compositions within the CTAB-SOS-H2O system. The significant advantages of this innovative approach include (1) the use of a single technique to demarcate distinct aggregates of the ternary system, (2) the application of a more generally accessible instrumental technique to the study of aggregate behavior, and (3) the utilization of a technique with the sensitivity to potentially both identify the aggregates present and characterize aggregate structure and morphology. In particular, given the aggregate-dependent variation in prodan spectral shape, one might initially seek a single measure, such as the ratio of observed fluorescence intensities at two emission wavelengths, as a parameter that would be sensitive to the nature of the aggregates present. However, we demonstrate that delineations of those aggregates with more subtle differences in overall prodan fluorescence spectral shape can be achieved by deconvoluting the spectra into overlapping Gaussian curves. Furthermore, while some characteristics of aggregates require substantial time to achieve an equilibrium value, for example, size and polydispersity,4 we show that the identity of the aggregate can be determined using fluorescence data acquired only a few days following sample preparation. With both the reduction in time arising from the use of a single analytical technique and the shortened time following sample preparation before analysis, a fluorescence approach is an effective means of delineating the microstructures of aqueous binary mixtures of surfactants. Experimental Section Materials. Aqueous stock solutions of the following surfactants were prepared: CTAB (Calbiochem, 99%) and SOS (Aldrich, 95%). All solutions were prepared with 18.2 MΩ ultrapure water obtained from a Milli-Q Millipore water filtration system (pH 5.0). An aqueous stock solution of prodan (Molecular Probes, Eugene, OR) was prepared, yielding a final concentration of 1.00 µM prodan when added to the surfactant solutions. The extreme spectral sensitivity of prodan enables the use of low concentrations to ensure that the probe does not alter phase behavior.44,45 Low concentrations also minimize the concentration of “free” prodan within a sample (i.e., prodan not interacting with surfactant aggregates), thereby enhancing our ability to resolve the presence of prodan incorporated within aggregates. (47) Karukstis, K. K.; Zieleniuk, C. A.; Fox, M. J. Langmuir 2003, 19, 10054-10060. (48) Karukstis, K. K.; Thonstad, S. C.; Hall, M. E. J. Dispersion Sci. Technol. 2002, 23, 737-744.

Karukstis et al. We varied the prodan concentration to find an optimal level (1 µM) that reduces the contribution of the 520-530-nm component (free prodan)45,46 while generating measurable fluorescence intensities reflecting prodan incorporation within surfactant aggregates. For those phase regions where vesicles are present, we have compared the average hydrodynamic radii and sample polydispersity obtained from light scattering measurements in both the presence and absence of prodan to ensure that the probe does not alter vesicle size or size distribution. Fluorescence Measurements and Analyses. A PerkinElmer LS-50B fluorescence spectrophotometer was used to obtain fluorescence emission spectra at 25 °C using an excitation wavelength of 340 nm. The emission wavelength was varied from 350 to 600 nm, while the excitation and emission slit widths were kept constant at 2.5 nm. Fluorescence spectra were recorded immediately after sample mixing and 1, 3, and 5 days after sample preparation; the results presented are for spectra recorded 5 days after sample preparation. Generally, we observed little change in spectral shape and intensity between the 1- and 3-day measurements. After conversion of wavelength to frequency, spectra were analyzed by deconvolution into overlapping Gaussian curves with the use of a commercial nonlinear least-squaresfitting method (PeakFit, SPSS Science). An iterative MarquardtLevenberg fitting algorithm49,50 was used to obtain the minimum number of reproducible fluorescing components using the adjustable parameters of the center, width, and amplitude of each Gaussian curve. The percent area of each Gaussian curve is also characterized. As with any fit involving multiple adjustable parameters, a unique solution is desirable but often not possible as multiple statistically equivalent minima may exist. Multiple attempts to fit the data with different starting parameters can generally provide a survey of the extent of statistically equivalent parameter sets.51 Comparing several deconvolutions of an overall spectrum, a “good” fit is then judged by several criteria including a minimum in the goodness-of-fit parameter χ2 (the weighted sum of the squares of the deviations),51 a random residuals plot with no systematic features,52 and a maximum value for the square of the multiple correlation coefficient, r2.51 From these statistically acceptable fits, a good fit is further judged by the reproducibility in the values for the centers of the Gaussian curves, that is, emission λmax values, for solution compositions ascribed to the same phase region. While distinct deconvolutions among all phase regions cannot be guaranteed, the spectral sensitivity of prodan enhances a successful outcome. Light Scattering Measurements. Multiangle light scattering measurements were made with a Wyatt Technology DAWN EOS (Enhanced Optical System) instrument poised at 25 °C. The Wyatt light scattering instrument has an 18-angle lightscattering geometry coupled with a 30 mW diode laser for simultaneous multiangle detection of light scattering with temperature control. Vesicle hydrodynamic radii in the range of 15-500 nm can be determined using Wyatt Technology’s ASTRA software based on the Debye method. Light scattering measurements were made 1 day, 4 days, 11 days, and 3 weeks after sample preparation; all results are presented and indicate that cationic-rich vesicles vary little in size between 4 and 11 days, while anionic-rich vesicles, particularly at higher total surfactant concentrations, grow 20-40% in size during the 3-week period.

Results Fluorescence Analysis Using Intensities at Fixed Wavelengths. In the dilute surfactant region of the CTAB-SOS-H2O system, up to nine solution-phase aggregate regions occur at fixed total surfactant concentration, including aggregates of CTAB-rich rodlike micelles (R), CTAB-rich rodlike micelles and cationic-rich unilamellar vesicles (R + VC), unilamellar cationic-rich vesicles (VC), anionic-rich unilamellar vesicles and a lamellar phase (49) Marquardt, D. W. J. Soc. Ind. Appl. Math. 1963, 11, 431-441. (50) Levenberg, K. Q. Appl. Math. 1944, 2, 164-168. (51) Bevington, P. R.; Robinson, D. K. Data Reduction and Error Analysis for the Physical Sciences, 2nd ed.; WCB/McGraw-Hill: Boston, 1992; Chapter 8. (52) Luokkala, B. B.; Garoff, S.; Suter, R. M. Phys. Rev. E 2000, 62, 2405-2415.

Delineation of Surfactant Microstructures

Langmuir, Vol. 20, No. 1, 2004 67

Figure 3. The ratio of prodan fluorescence intensities at 430 and 485 nm (I430/I485) in CTAB-SOS-H2O samples prepared at total surfactant concentrations of 1.0, 1.5, 2.0, 2.5, 3.0, and 4.0 wt %. Each point represents a sample with an increasing SOS composition. The dotted vertical lines demarcate the boundaries between regions with different combinations of surfactant aggregates as reported in ref 4.

(V + L), unilamellar anionic-rich vesicles (V), rodlike SOSrich micelles and unilamellar anionic-rich vesicles (R + V), SOS-rich rodlike micelles (R), and spherical SOS-rich micelles (M). An unresolved multiphase region is also located on the SOS-rich side of the phase diagram (I). Samples were prepared at varying CTAB and SOS proportions such that the total surfactant concentration was fixed at 1.0, 1.5, 2.0, 2.5, 3.0, and 4.0 wt %. Compositions were selected to target the various solutionphase aggregate regions exhibited by the ternary system. At [total surfactant] ) 1.0 wt %, the sequence of samples prepared with increasing SOS wt % aimed to prepare the following aggregates: R f R + VC f VC f V + L f V f M. For [total surfactant] ) 1.5, 2.0, 2.5, and 3.0 wt %, the intended aggregates were R f R + VC f V + L f V f M. The sequence of aggregates for 4.0 wt % was designed to be R f R + VC f V + L f I f R + V f M. Prodan (1 µM) fluorescence spectra were recorded, and ratios of fluorescence intensities observed at pairs of wavelengths between 430 and 500 nm were analyzed to find a combination yielding ratios with the greatest distinction among the various aggregate regions. A ratio of fluorescence intensities at 430 and 485 nm (I430/I485) was selected based on this criterion. Figure 3 presents these ratios for each total surfactant concentration, with the dotted vertical lines indicating the intended demarcations between aggregate regions. The combined data in Figure 4 as a function of SOS wt % show the ratios observed by aggregate region. The CTAB-rich (2) and SOS-rich ([) micellar regions show intensity ratios between about 0.1 and 0.2 (with the lowest values of pure micelles of a single surfactant); the regions containing both cationic-rich micelles and vesicles (b) exhibit ratios around 0.3-0.4; both CTAB- and SOS-rich vesicles (1) yield fluorescence intensity ratios around about 0.40 and 0.44; and the SOSrich vesicle and lamellar samples (9) exhibit the highest ratios, greater than or equal to about 0.45. The ratio for the sample in the unresolved multiphase region is intermediate between that of vesicle samples and vesicle + lamellae samples. Using these ranges of fluorescence

Figure 4. The ratio of prodan fluorescence intensities at 430 and 485 nm (I430/I485) in CTAB-SOS-H2O samples prepared at total surfactant concentrations of 1.0, 1.5, 2.0, 2.5, 3.0, and 4.0 wt %. Each point represents a distinct sample with an increasing SOS composition. Samples expected4 to contain similar surfactant aggregates are delineated with the same symbol, including aggregates of CTAB and CTAB-rich rodlike micelles (R) (2), CTAB-rich rodlike micelles and cationic-rich unilamellar vesicles (R + VC) (b), unilamellar cationic-rich vesicles or unilamellar anionic-rich vesicles (VC or V) (1), anionic-rich unilamellar vesicles and a lamellar phase (V + L) (9), and SOS-rich rodlike micelles (R) and spherical SOS-rich micelles (M) ([). Aggregate regions represented by only a single sample (i.e., rodlike SOS-rich micelles and unilamellar anionicrich vesicles (R + V) or the SOS-rich unresolved multiphase region (I) are not denoted in the figure. The open symbol 4 represents samples that are believed to be present in the region corresponding to CTAB-rich rodlike micelles (R) rather than the region of CTAB-rich rodlike micelles and cationic-rich unilamellar vesicles (R + VC). The open symbol 0 corresponds to a sample that is believed to be in the region containing anionicrich unilamellar vesicles and a lamellar phase (V + L) rather than the region of CTAB-rich rodlike micelles and cationic-rich unilamellar vesicles (R + VC).

ratios, only four samples appear to be located in slightly different aggregate regions. Sample 3 at [total surfactant]

68

Langmuir, Vol. 20, No. 1, 2004

Karukstis et al.

Figure 5. The normalized fluorescence emission spectra of prodan (1 µM) in CTAB-SOS-H2O samples with [total surfactant] ) 1.5 wt %. The assignment of samples to particular surfactant aggregates is given in the text.

) 1.0 wt % and sample 4 at both 1.5 and 2.5 wt % (all denoted 4 in Figure 4) appear to be in the CTAB-rich rodlike micellar region R rather than the biaggregate region of R + VC; sample 5 at [total surfactant] ) 3.0 wt % (denoted 0) appears to be in the SOS-rich vesicle and lamellar region (V + L) rather than in the cationic-rich micelle and vesicle region (R + VC). Fluorescence Analysis Using Deconvoluted Spectra. Samples with [CTAB] + [SOS] ) 1.5 wt %. Thirteen samples were prepared of varying CTAB and SOS proportions such that the total surfactant concentration was fixed at 1.5 wt %. Five aggregate regions occur at this fixed total surfactant concentration. Sample 1 was a pure CTAB micellar solution. Samples 2 and 3 were chosen in the rodlike CTAB-rich micellar region (R), samples 4-6 in the biaggregate region of CTAB-rich rodlike micelles and liquid crystalline cationic-rich vesicles (R + VC), samples 7-9 in the SOS-rich vesicle and lamellar region (V + L), samples 10-12 in the SOS-rich vesicle region (V), and sample 13 in the narrow SOS-rich spherical micellar region (M). Figure 5 presents the normalized fluorescence spectra of these samples with added 1 µM prodan. The particular grouping of spectra in each diagram was selected to emphasize spectral similarities and differences. Three sets of samples yielded identical normalized spectra for each sample within the set (distinct from spectra in other sets): samples 2-4, 7-9, and 10-12. Sample 13 gave rise to an extremely separate spectrum. The spectrum for sample 1 was slightly red-shifted from those observed for samples 2-4. The spectra for samples 5 and 6 were intermediate between those exhibited for samples 2-4 and samples 7-9. The parameters of the Gaussian curves resulting from spectral deconvolution are presented in Table 1. Each spectrum was deconvoluted into a total of 2-3 Gaussianshaped curves attributed to prodan associated with surfactant aggregates and one minor Gaussian associated with free prodan. For prodan associated with aggregates,

the Gaussians were centered at λ1 between 483 and 512 nm, λ2 between 447 and 465 nm, and λ3 between 418 and 436 nm. The minor Gaussian reflecting free prodan (not shown) occurred at 520-530 nm, consistent with the λmax value observed for prodan in water.45,46 A sample deconvolution is illustrated in Figure 6. These data further emphasize the spectral shape similarities and differences. Sample 1 was distinguished from samples 2-4 by the λmax of Gaussian 2, 465 nm (sample 1) versus 455 nm (average of samples 2-4). Compared with samples 1-4, samples 5 and 6 exhibited a shorter λ1 (average of 491 nm vs 485 nm, respectively) and a greater contribution of Gaussian 2 relative to Gaussian 1 (average ratio of areas of 0.32 vs 0.24, respectively). Each of these factors contributes to the blue-shift of spectra 5 and 6 relative to spectra 1-4. Samples 7-9 exhibit the largest contribution of Gaussian 2 relative to Gaussian 1, averaging a ratio of Gaussian areas equal to 0.40. The slight spectral red-shift in samples 10-12 from the position of spectra 7-9 arises primarily from the decrease in the contribution of Gaussian 2 for samples 10-12 and the larger value of λ2 for samples 10-12. Samples with [CTAB] + [SOS] ) 2.5 wt % by Weight. Fifteen samples were prepared of varying CTAB and SOS proportions such that the total surfactant concentration was fixed at 2.5 wt %. The same five aggregate regions as noted above occur at this fixed total surfactant concentration. Sample 1 was a pure CTAB micellar solution. Samples 2 and 3 were chosen in the rodlike CTAB-rich micellar region (R), samples 4-6 in the biaggregate region of CTAB-rich rodlike micelles and liquid crystalline cationic-rich vesicles (R + VC), samples 7-9 in the SOSrich vesicle and lamellar region (V + L), samples 10-12 in the SOS-rich vesicle region (V), and samples 13-15 in the SOS-rich spherical micellar region (M). The normalized fluorescence spectra of these samples with added 1 µM prodan were recorded and are presented in Figure 7. The parameters of the Gaussian curves resulting from spectral

Delineation of Surfactant Microstructures

Langmuir, Vol. 20, No. 1, 2004 69

Table 1. Prodan Fluorescence Parameters of CTAB-SOS-Water Samples at [Total Surfactant] ) 1.5 wt % intended aggregatesa sample phase(s) 1 2 3 4 5 6 7 8 9 10 11 12 13

R R R R + VC R + VC R + VC V+L V+L V+L V V V M

composition,b % CTAB

flr int ratio,c I430/I485

λ1/nm

30.0 28.0 26.0 24.0 22.0 20.0 18.0 16.0 14.0 11.0 7.0 3.0 0.0

0.111 0.149 0.160 0.195 0.324 0.399 0.460 0.464 0.439 0.419 0.401 0.395 0.136

494 491 490 490 483 487 488 488 489 489 487 488 512

major Gaussians reflecting prodan incorporationd λ2/nm λ3/nm A2/A1e A3/A1f 465 456 455 454 449 452 447 451 450 452 453 453

416 426 418 421 416 420 420 422 426 425 436

0.24 0.26 0.27 0.13 0.30 0.33 0.41 0.40 0.39 0.34 0.29 0.31 0

0 0 0.02 0.09 0.08 0.14 0.07 0.15 0.15 0.15 0.21 0.20 0.09

a Aggregate assignments based on interpretation of compositions in the ternary phase diagram in ref 4. b All percents given are wt % with %H2O ) 70.0% and %SOS ) 30.0% - %CTAB. c The ratio is calculated using the fluorescence emission intensities at 430.0 and 485.0 nm; uncertainty of (0.005. d The centers of the Gaussian curves resulting from the spectral deconvolution are reported; uncertainty of (2 nm. e The ratio of the areas of the Gaussian peaks centered at λ2 and λ1; uncertainty of (0.01. f The ratio of the areas of the Gaussian peaks centered at λ3 and λ1; uncertainty of (0.01.

after sample preparation. Vesicle radii were monitored 1 day, 4 days, 11 days, and 3 weeks following sample preparation. While cationic-rich vesicles appeared to reach their equilibrium size after about 4 days, anionic-rich vesicles continued to grow 20-40% in size over the 3-week period. In contrast, for either cationic-rich or anionic-rich vesicles, all normalized fluorescence spectra, independent of sample composition or timing following sample preparation, were superimposable (results not shown). The Gaussian deconvolution data in Table 3 emphasize such similar spectral shape. Discussion

Figure 6. Deconvolution of the prodan emission spectrum for a CTAB-SOS-H2O sample in the vesicle region, yielding four Gaussian components centered at approximately 430, 455, 490, and 530 nm.

deconvolution are summarized in Table 2. These data support the spectral shape similarities and differences observed for the 1.5 wt % samples. For example, the pure CTAB sample 1 was distinguished from samples 2-4 by the λmax of Gaussian 2, 459 nm (sample 1) versus 451 nm (average of samples 2-4). Compared with samples 1-4, samples 5 and 6 exhibit a substantially greater contribution of Gaussian 2 relative to Gaussian 1 (average ratio of areas of 0.54 vs 0.13, respectively). Samples 7-9 exhibit large contributions of Gaussian 2 relative to Gaussian 1, averaging a ratio of Gaussian areas equal to 0.45. The contribution of Gaussian 2 relative to Gaussian 1 decreases further for samples 10-12 (average of 0.37) and significantly more for samples 13 and 14 with no contribution of Gaussian 2 for sample 15. The value of λ1 for samples 13-15 also shows an increase relative to the wavelength observed for samples 2-12. CTAB-Rich and SOS-Rich Vesicles. To further demonstrate the utility of the fluorescence approach to surfactant aggregate characterization, we prepared samples within the cationic-rich and anionic-rich vesicle regions. Table 3 summarizes the compositions of the samples, with cationic-rich vesicles prepared at a constant ratio of wt % CTAB/wt % SOS ) 2.1:1.0 and with anionicrich vesicles prepared at a constant ratio of wt % CTAB/ wt % SOS ) 0.5:1.0. Prodan fluorescence emission spectra were recorded 1 day (not shown) and 5 days (Figure 8)

The fluorescence signature of prodan in surfactant aggregates is a composite spectrum with contributions from the probe population within various microdomains of the aggregate. A Gaussian fluorescence component centered at 493-509 nm has been attributed to prodan situated at a micellar surface via weak dipole interactions with the surfactant headgroups.42,44,45 The wavelength observed depends on the net micellar surface charge. The small shift in emission λmax from that in aqueous solution (λmax ) 522 nm45,46) is consistent with the assignment of prodan within the hydration sphere of the micellar aggregate. A similar fluorescence component between 486 and 506 nm is ascribed to prodan distributed at the watersurfactant interface in reverse micelles,43-45 cationic bilayers,45 and unilamellar vesicles with varying net surface charges.45,47 A 464-466-nm component observed in cationic micelles42,44,45 is consistent with emission from prodan residing in a region that is more nonpolar than the micellar surface but not as hydrophobic as the micellar core. This fluorescence component is ascribed to a strong interaction of the aromatic probe at the cationic surfactant headgroups. The presence of a similar component (458-468 nm) in other cationic systems (zwitterionic micelles at low pH where the protonated surfactant has cationic character),44,45 reverse micelles,43-45 unilamellar vesicles,45,47 and bilayers45) corroborates the assignment of this component to an interaction of probe and cationic headgroup. A 425-442-nm component resolved in nonionic, cationic, and zwitterionic micelles42,44,45 is attributed to prodan molecules incorporated via hydrophobic forces within the micellar core. The significant blue-shift in the prodan emission λmax is indicative of this nonpolar microregion with limited water accessibility. In analogy, an emission λmax between 425 and 442 nm is assigned to prodan in the

70

Langmuir, Vol. 20, No. 1, 2004

Karukstis et al.

Figure 7. The normalized fluorescence emission spectra of prodan (1 µM) in CTAB-SOS-H2O samples with [total surfactant] ) 2.5 wt %. The assignment of samples to particular surfactant aggregates is given in the text. Table 2. Prodan Fluorescence Parameters of CTAB-SOS-Water Samples at [Total Surfactant] ) 2.5 wt % intended aggregatesa sample phase(s) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

R R R R + VC R + VC R + VC V+L V+L V+L V V V M M M

composition,b % CTAB

flr int ratio,c I430/I485

λ1/nm

50.0 46.0 43.0 40.0 36.0 33.0 30.0 26.0 22.0 19.0 14.0 9.0 6.0 3.0 0.0

0.117 0.151 0.177 0.230 0.391 0.424 0.458 0.458 0.444 0.430 0.404 0.212 0.199 0.172 0.144

494 490 488 487 490 490 489 489 491 490 489 492 498 499 510

major Gaussians reflecting prodan incorporationd λ2/nm λ3/nm A2/A1e A3/A1f 459 451 451 452 454 453 450 454 452 452 452 454 451 457

420 424 427 416 418 418 423 420 421 422 423 419 429 440

0.18 0.13 0.12 0.08 0.56 0.52 0.47 0.44 0.44 0.37 0.36 0.39 0.16 0.09 0

0 0.04 0.06 0.09 0.08 0.14 0.13 0.21 0.13 0.14 0.16 0.15 0.06 0.11 0.07

a Aggregate assignments based on interpretation of compositions in the ternary phase diagram in ref 4. b All percents given are wt % with %H2O ) 50.0% and %SOS ) 50.0% - %CTAB. c The ratio is calculated using the fluorescence emission intensities at 430.0 and 485.0 nm; uncertainty of (0.005. d The centers of the Gaussian curves resulting from the spectral deconvolution are reported; uncertainty of (2 nm. e The ratio of the areas of the Gaussian peaks centered at λ2 and λ1; uncertainty of (0.01. f The ratio of the areas of the Gaussian peaks centered at λ3 and λ1; uncertainty of (0.01.

surfactant interfacial region of reverse micelles,43-45 singlesurfactant and binary-surfactant unilamellar vesicles,45,47 and single-surfactant bilayer systems.45 With these similar, yet distinctive, emission wavelengths for prodan populations within surfactant aggregates, the described fluorescence analysis of the CTAB-SOS-H2O system does provide an approach for delineating aggregate regions. The overall fluorescence spectral shape and the ratio of fluorescence intensities at two selected wavelengths, 430 and 485 nm, provide two means of delineating the microstructures present. In particular, several significant observations can be made from the spectral analysis: (1) The transitions between aggregate regions are marked for R f R + VC and V f M, gradual for R + VC f V + L, and slight but detectable

for V + L f V, supporting our claim that a fluorescence approach to delineation of the aggregate regions in such surfactant systems is a viable method. (2) Biaggregate regions R + VC and V + L exhibit more scattering as reflected by the raised spectral baseline, suggesting a preliminary means of assigning multiaggregate regions. (3) Spectra from samples within the single aggregate regions R (i.e., mixed CTAB-SOS rodlike micelles) and V (SOS-rich vesicles) exhibit superimposable spectra. Such an observation indicates little variation in the distribution of prodan within the aggregates of a given phase region. This fluorescence characteristic provides a means of identifying aggregates that are comparable in size and structure (e.g., vesicles of similar size and similar net outer surface charge). (4) The pure CTAB and mixed CTAB-

Delineation of Surfactant Microstructures

Langmuir, Vol. 20, No. 1, 2004 71

Table 3. Fluorescence and Light Scattering Characterization of CTAB-Rich and SOS-Rich Vesicles major Gaussians reflecting prodan incorporationb sample

wt % H2O

λ1/nm

λ2/nm

λ3/nm

c

A2/A1

average vesicle radius/nm d

A3/A1

after 1 day

1 2 3 4

88.0 83.8 80.4 74.1

486 489 489 485

451 452 452 450

CTAB-Rich Vesiclesa (wt % CTAB/wt % SOS ) 2.1) 421 0.38 0.14 119 ( 2 421 0.40 0.15 104 ( 2 418 0.41 0.10 92 ( 3 418 0.42 0.07 89 ( 3

1 2 3 4 5 6 7 8 9

86.5 80.7 72.2 66.5 60.8 52.2 43.7 38.0 29.4

488 492 489 490 490 490 490 491 490

452 454 452 452 452 452 451 452 451

SOS-Rich Vesiclesa (wt % CTAB/wt % SOS ) 0.50) 423 0.37 0.17 138 ( 2 422 0.37 0.13 75 ( 1 422 0.35 0.16 66 ( 1 423 0.37 0.16 58 ( 1 422 0.35 0.16 83 ( 2 421 0.36 0.15 68 ( 1 420 0.36 0.13 56 ( 1 421 0.37 0.12 57 ( 1 420 0.37 0.11 53 ( 1

after 4 days

after 11 days

after 3 wks

127 ( 2 114 ( 2 118 ( 1 108 ( 1

128 ( 2 118 ( 2 121 ( 1 110 ( 1

121 ( 1 120 ( 2 122 ( 1 109 ( 1

136 ( 2 76 ( 1 68 ( 1 67 ( 1 83 ( 2 72 ( 1 76 ( 1 69 ( 1 62 ( 1

137 ( 2 81 ( 1 73 ( 1 75 ( 1 88 ( 4 81 ( 1 84 ( 1 76 ( 1 70 ( 1

135 ( 2 90 ( 2 75 ( 1 81 ( 1 100 ( 3 94 ( 2 88 ( 1 78 ( 2 74 ( 1

a Aggregate assignments as vesicles based on interpretation of compositions in the ternary phase diagram in ref 4. b Fluorescence spectra were recorded 5 days after sample preparation. The centers of the Gaussian curves resulting from the spectral deconvolution are reported; uncertainty of (2 nm. c The ratio of the areas of the Gaussian peaks centered at λ2 and λ1; uncertainty of (0.01. d The ratio of the areas of the Gaussian peaks centered at λ3 and λ1; uncertainty of (0.01.

Figure 8. Prodan fluorescence emission spectra in cationicrich and anionic-rich vesicle samples in Table 3. Spectra were recorded 5 days after sample preparation and are labeled with the sample number listed in Table 3.

SOS rodlike micelles are distinguished by a small spectral shift, providing a means of assessing the surfactant “purity” of an aggregate. A similar distinction between pure SOS and SOS-rich micelles is also noted. (5) The similar fluorescence spectra for samples 1, 2, and 3 at a total surfactant concentration of 1.0 wt % and samples 2, 3, and 4 at total surfactant concentrations of 1.5 and 2.5 wt % suggest that the earlier assignment4 of the latter composition to the biaggregate region of CTAB-rich rodlike micelles and vesicles may be in error. Furthermore, the similar fluorescence spectra of samples 5, 6, and 7 at a total surfactant concentration of 3.0 wt % suggest that the earlier assignment of composition 5 to the biaggregate region of CTAB-rich rodlike micelles and vesicles may also be in error. A “narrowing” of the range of compositions giving rise to this biaggregate region would accommodate our fluorescence observations. Thus, a fluorescence approach can be used to refine previously characterized surfactant systems. The deconvolution data in Tables 1 and 2 enable further refinements to aggregate region borders and allow interpretation of aggregate structure. For example, while the relative areas of the Gaussians observed for pure CTAB micelles and CTAB-rich micelles are quite similar, the values of λ1 and λ2 are distinctive for the pure and mixed micelles. The decrease in λ1 with added SOS reflects a

decrease in the water accessibility of the headgroup region as added negative charge enables positively charged headgroups to approach more closely, excluding some water. The decrease in λ2 is similar to that observed for other alkyltrimethylammonium bromide surfactants (CnTAB where n ) 12-18) as increasing surfactant concentrations change the micellar shape from spherical to rodlike.44,45 The increasing amounts of Gaussians 2 and 3 for those regions containing vesicles are consistent with the higher proportions of these two components observed in other vesicle systems.45,47 With added SOS, the more “neutral” outer vesicle surface enables prodan to penetrate into the hydrophobic bilayer. As CTAB content continues to decrease, the loss of the Gaussian at λ2 is consistent with the greater degree of electrostatic repulsion of the π electrons of the naphthalene ring of prodan with the negative SOS headgroup charge. While the use of molecular probes to characterize surfactant microstructures is an indirect approach, our investigations reveal the sensitivity of this method to the surfactant aggregates present in a given sample. Interestingly, the fluorescence approach enables the microstructure determination in a rather rapid and streamlined fashion as the fluorescence spectral shape distinguishing a particular aggregate region does not appear to require an equilibrium aggregate size to be achieved. Furthermore, as phase separation can often require substantially longer periods of time than is necessary for aggregate formation,4,53,54 the acquisition of fluorescence spectra without waiting for phase separation will also significantly accelerate phase diagram characterization. Conclusion This study demonstrates that a fluorescence spectroscopic approach using the probe prodan is a sensitive and efficient technique to demarcate aggregate regions in surfactant systems. With complementary light scattering and optical microscopy measurements, further characterizations of aggregate identity, structure, and morphology are possible. As there is considerable fundamental and commercial interest in characterizing surfactant mixtures due to the enhanced properties of these systems (53) Viseu, M. I.; Edwards, K.; Campos, C. S.; Costa, S. M. B. Langmuir 2000, 16, 2105-2114. (54) Viseu, M. I.; Velazquez, M. M.; Campos, C. S.; Garcia-Mateos, I.; Costa, S. M. B. Langmuir 2000, 16, 4882-4889.

72

Langmuir, Vol. 20, No. 1, 2004

relative to single surfactants, our spectroscopic method is a promising approach to assist in understanding the principles governing aggregate formation in complex surfactant systems. Acknowledgment. The project described was supported by Grant Number 2 R15 GM55911-02 from the National Institute of General Medical Sciences of the National Institutes of Health. Acknowledgment is also made to the donors of the Petroleum Research Fund,

Karukstis et al.

administered by the American Chemical Society, for the partial support of this research. This research was also supported in part by a grant from the National Science Foundation Research Experiences for Undergraduates Program (CHE-0097262). The authors also acknowledge the contributions of the preliminary investigations of Cecilia Giddings and Neel Joshi of Harvey Mudd College to this work. LA0351764