11458
Langmuir 2007, 23, 11458-11464
Micellization of Dissymmetric Cationic Gemini Surfactants and Their Interaction with Dimyristoylphosphatidylcholine Vesicles Yanru Fan,† Yajuan Li,† Meiwen Cao,† Jinben Wang,† Yilin Wang,*,† and Robert K. Thomas‡ Key Laboratory of Colloid and Interface Science, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100080, People’s Republic of China, and Physical and Theoretical Chemistry Laboratory, Oxford UniVersity, South Parks Road, Oxford OX1 3QZ, United Kingdom ReceiVed May 22, 2007. In Final Form: August 7, 2007 The micellization process of a series of dissymmetric cationic gemini surfactants [CmH2m+1(CH3)2N(CH2)6N(CH3)2C6H13]Br2 (designated as m-6-6 with m ) 12, 14, and 16) and their interaction with dimyristoylphosphatidylcholine (DMPC) vesicles have been investigated. In the micellization process of these gemini surfactants themselves, critical micelle concentration (cmc), micelle ionization degree, and enthalpies of micellization (∆Hmic) were determined, from which Gibbs free energies of micellization (∆Gmic) and entropy of micellization (∆Smic) were derived. These properties were found to be influenced significantly by the dissymmetry in the surfactant structures. The phase diagrams for the solubilization of DMPC vesicles by the gemini surfactants were constructed from calorimetric results combining with the results of turbidity and dynamic light scattering. The effective surfactant to lipid ratios in the mixed aggregates at saturation (Resat) and solubilization (Resol) were derived. For the solubilization of DMPC vesicles, symmetric 12-6-12 is more effective than corresponding single-chain surfactant DTAB, whereas the dissymmetric m-6-6 series are more effective than symmetric 12-6-12, and 16-6-6 is the most effective. The chain length mismatch between DMPC and the gemini surfactants may be responsible for the different Re values. The transfer enthalpy per mole of surfactant within the coexistence range may be associated with the total hydrophobicity of the alkyl chains of gemini surfactants. The transfer enthalpies of surfactant from micelles to bilayers are always endothermic due to the dehydration of headgroups and the disordering of lipid acyl chain packing during the vesicle solubilization.
Introduction
* To whom correspondence should be addressed. E-mail: yilinwang@ iccas.ac.cn. † Chinese Academy of Sciences. ‡ Oxford University.
Surfactants are indispensable reagents in the solubilization and reconstitution of lipid membranes. The structure and topology of surfactant molecules play an important role on the solubilization of lipid vesicles. Among the studies on surfactant-lipid interaction, nonionic surfactants have received much attention because of their mild effect on membrane proteins.13-16 Edwards and Almgren14 have studied the effect of surfactant headgroup size on the solubilization of lecithin vesicles for C12En series and found that, for the surfactant with a larger headgroup, lower surfactant concentration is needed to induce the structural transition. In the case of ionic surfactant-induced lipid solubilization, physiological surfactant bile salts which have a hydrophobic steroidal backbone with one to three hydroxyl groups have been widely investigated.17-19 Majhi and Blume10 found that bile salts are more effective than classical head-tail surfactant sodium dodecyl sulfate (SDS). In addition, the solubilization of lipid vesicles by a series of alkyl sulfate surfactants has been investigated by cryo-transmission electron microscopy (cryoTEM) and light scattering.4 These results show that surfactant chain length has a profound influence on both the surfactant amount needed for the solubilization of lipid bilayers and the type of structures formed during the vesicle to micelle transition.
(1) Heerklotz, H.; Seelig, J. Biochim. Biophys. Acta 2000, 1508, 69. (2) Almgren, M. Biochim. Biophys. Acta 2000, 1508, 146. (3) Lichtenberg, D.; Opatowski, E.; Kozlov, M. M. Biochim. Biophys. Acta 2000, 1508, 1. (4) Silvander, M.; Karlsson, G.; Edwards, K. J. Colloid Interface Sci. 1996, 179, 104. (5) Tan, A.; Ziegler, A.; Steinbauer, B.; Seelig, J. Biophys. J. 2002, 83, 1547. (6) Pata, V.; Ahmed, F.; Discher, D. E.; Dan, N. Langmuir 2004, 20, 3888. (7) Levy, D.; Gulik, A.; Seigneuret, M.; Rigaud, J. Biochemistry 1990, 29, 9480. (8) Lichtenberg, D.; Robson, R. J.; Dennis, E. A. Biochim. Biophys. Acta 1983, 737, 285. (9) Lichtenberg, D. Biochim. Biophys. Acta 1985, 821, 470. (10) Majhi, P. R.; Blume, A. J. Phys. Chem. B 2002, 106, 10753. (11) Keller, M.; Kerth, A.; Blume, A. Biochim. Biophys. Acta 1997, 1326, 178.
(12) Lo´pez, O.; Co´cera, M.; Coderch, L.; Parra, J. L.; Barsukov, L.; De la Maza, A. J. Phys. Chem. B 2001, 105, 9879. (13) Schnitzer, E.; Lichtenberg, D.; Kozlov, M. M. Chem. Phys. Lipids 2003, 126, 55. (14) Edwards, K.; Almgren, M. Langmuir 1992, 8, 824. (15) Edwards, K.; Almgren, M.; Bellare, J.; Brown, W. Langmuir 1989, 5, 473. (16) Johnsson, M.; Edwards, K. Langmuir 2000, 16, 8632. (17) Hildebrand, A.; Beyer, K.; Neubert, R.; Garidel, P.; Blume, A. J. Colloid Interface Sci. 2004, 279, 559. (18) Hildebrand, A.; Neubert, R.; Garidel, P.; Blume, A. Langmuir 2002, 18, 2836. (19) Mazer, N. A.; Benedek, G. B.; Carey, M. C. Biochemistry 1980, 19, 601.
Numerous studies have been performed on surfactant-induced lipid solubilization which leads to vesicle-micelle transformation by incorporation of surfactants into lipid vesicles.1-5 These studies have been driven by the need to solubilize biomembranes for the isolation, purification, reconstitution, and crystallization of membrane proteins or for the incorporation of pharmacological agents into drug-carrying vesicles.6,7 The mechanism of this kind of vesicle-micelle transition has been described by Lichtenberg using a three-stage model.8,9 Initially surfactants incorporate into vesicle bilayers up to a critical surfactant/lipid ratio Resat. Beyond this value, lipid saturated mixed micelles coexist with surfactant saturated vesicles. When the added surfactant concentration exceeds a critical value Resol, the solubilization of mixed vesicles into mixed micelles is completed. This model has been extensively used to explain the phenomena observed in the solubilization of lipid vesicles.10-12
10.1021/la701493s CCC: $37.00 © 2007 American Chemical Society Published on Web 10/05/2007
Micellization of Cationic Gemini Surfactants
In recent years, gemini surfactants with two hydrophobic chains and two hydrophilic groups connected by a spacer group have attracted much interest, because they often have superior properties compared to monomeric surfactants such as much lower critical micelle concentration (cmc), greater surface activities, better wetting properties, and so on.20-24 Besides a good deal of studies on symmetric gemini surfactants, considerable attention has also been paid to dissymmetric gemini surfactants with two dissymmetric hydrophobic chains. This dissymmetry in the hydrophobic chains influences the diversity of phase behavior and thermodynamic parameters notably.25-30 Although the special structure of gemini surfactants has led to their unique properties on micellization, to what extent it will affect their interaction with lipid vesicles has not been welldocumented. In the literature several papers have related to the gemini surfactant/lipid systems. The effect of gemini surfactants on the PC bilayer thickness in the fluid lamellar phase has been studied using X-ray diffraction.31 The properties of the mixed micelles at higher gemini surfactant/lipid molar ratio have been characterized using fluorescence spectroscopy by Bakshi et al.32,33 However, all these studies have only focused on one special aspect of gemini surfactant-lipid interaction while not involving an investigation of the solubilization process. Therefore, detailed and systematic study on the gemini surfactant-induced solubilization of lipid vesicles is necessary. In the present study, the solubilization process of DMPC vesicles induced by gemini surfactants were investigated by isothermal titration microcalorimetry (ITC), turbidity, and dynamic light scattering (DLS). One symmetric and three dissymmetric gemini surfactants have been employed for comparison. They have the structure of [CmH2m+1(CH3)2N(CH2)6N(CH3)2CnH2n+1]Br2, designated as m-6-6 with m ) 12, 14, and 16 and n ) 6 for dissymmetric ones and 12-6-12 with m ) n ) 12 for the symmetric one. Because the dissymmetric m-6-6 series have never been reported, we studied their micellization process first. The second part is focused on their interaction with DMPC vesicles. The main purpose of the present work include two sides: One side is to elucidate whether gemini surfactants behave in a way similar to that of single-chain surfactants when interacting with phospholipid vesicles, and another side is to find out the effect of dissymmetry and chain length of the hydrophobic chains of gemini surfactants on their solubilization ability to DMPC vesicles.
Langmuir, Vol. 23, No. 23, 2007 11459 (DMPC) was purchased from Sigma and used without further purification. Triply distilled water was used in all experiments. Preparation of Vesicles. An appropriate amount of DMPC was dissolved in chloroform and then was dried under vacuum on a rotary evaporator to obtain a thin film on the bottom of the flask. The dried film was subsequently suspended in a certain volume of triply distilled water to a final desired concentration and vortexed at the required temperature above Tc (gel to liquid crystal transition temperature) for several minutes. After that, the crude phospholipid suspension was sonicated for about 30 min in a water bath at 40 °C. These freshly prepared vesicle dispersions were then used directly for the solubilization experiments. Electrical Conductivity. This method was used to determine the cmc and micelle ionization degree (R) of the surfactants. The conductivity of the solutions was measured at 298.15 ( 0.1 K as a function of surfactant concentration using a JENWAY model 4320 conductivity meter. Concentrated surfactant solution was successively added to 8 mL of water, and sufficient time was allotted after each addition to allow the system to equilibrate. Isothermal Titration Calorimetry. The calorimetric measurements were taken in a TAM 2277-201 microcalorimetric system (Thermometric AB, Ja¨rfa¨lla, Sweden) with a stainless steel sample cell of 1 mL at 298.15 ( 0.01 K. The reference cell was filled with water. In the case of micellization experiments, the sample cell was initially loaded with 0.6 mL of water. Aliquots of concentrated gemini surfactant solutions of 10 µL were injected into the stirred sample cell using a 500 µL Hamilton syringe controlled by a Thermometric 612 Lund pump. In the case of solubilization experiment, the cell was initially loaded with 0.7 mL of DMPC vesicle dispersion, and aliquots of concentrated gemini surfactant solutions of 4 µL were injected into the stirred sample cell. The addition of surfactant solution and the measurements of heat were done as programmed. Turbidity Measurements. Concentrated surfactant solution was added in a stepwise manner to 1 mL vesicle dispersion. After each addition the sample was stirred for 10 min; thereafter, the absorbance was measured using a Hitachi UV-vis spectrophotometer (model U-4500). All the measurements were conducted at 298.15 ( 0.5 K. Dynamic Light Scattering. Measurements were carried out at 298.15 ( 0.5 K with a LLS spectrometer (ALV/SP-125) which employs a multi-τ digital time correlator (ALV-5000). A solid-state He-Ne laser (output power ) 22 mW at λ ) 632.8 nm) was used as a light source. The samples were introduced into a 7 mL glass bottle through a 0.45 µm Millipore filter prior to measurements. Thereafter DLS measurements were performed at a scattering angle of 90°. The autocorrelation function of scattering data was analyzed via the CONTIN method.
Experimental Section Materials. The gemini surfactants 12-6-12, 12-6-6, 14-6-6, and 16-6-6 were synthesized and purified according to the literature.34 The phospholipid 1,2-dimyristoyl-sn-glycero-3- phosphotidylcholine (20) Menger, F. M.; Littau, C. A. J. Am. Chem. Soc. 1993, 115, 10083. (21) Zana, R.; Benrroau, M.; Rueff, R. Langmuir 1991, 7, 1072. (22) Song, Li. D.; Rosen, M. J. Langmuir 1996, 12, 1149. (23) Menger, F. M.; Keiper, J. S.; Azov, V. Langmuir 2000, 16, 2062. (24) Li, Y. J.; Li, P. X.; Wang, J. B.; Wang, Y. L.; Yan, H. K.; Thomas, R. K. Langmuir 2005, 21, 6703. (25) Huc, I.; Oda, R. Chem. Commun. (Cambridge) 1999, 2025. (26) Oda, R.; Huc, I.; Candau, S. J. Chem. Commun. (Cambridge) 1997, 2105. (27) Oda, R.; Huc, I.; Homo, J. C.; Heinrich, B.; Schmutz, M.; Candau, S. Langmuir 1999, 15, 2384. (28) Sikiric´, M.; Primozˇicˇ, I.; Talmon, Y.; Filipovic´-Vincekovic´, N. J. Colloid Interface Sci. 2005, 281, 473. (29) Sikiric´, M.; Primozˇicˇ, I.; Filipovic´-Vincekovic´, N. J. Colloid Interface Sci. 2002, 250, 221. (30) Wang, X. Y.; Wang, J. B.; Wang, Y. L.; Ye, J. P.; Yan, H. K.; Thomas, R. K. J. Phys. Chem. B 2003, 107, 11428. (31) Dubnicˇkova´, M.; Yaradaikin, S.; Lacko, I.; Devı´nsky, F.; Gordeliy, V.; Balgavy´, P. Colliods Surf., B 2004, 34, 161. (32) Bakshi, M. S.; Singh, K.; Singh, J. J. Colloid Interface Sci. 2006, 297, 284. (33) Bakshi, M. S.; Singh, J.; Kaur, G. Chem. Phys. Lipids 2005, 138, 81. (34) Bai, G. Y.; Wang, J. B.; Wang, Y. L.; Yan, H. K.; Thomas, R. K. J. Phys. Chem. B 2002, 107, 6614.
Results Micellization of Dissymmetric Gemini Surfactants. Conductivity and ITC were employed to investigate the micellization process of these dissymmetric gemini surfactants. Figure 1 presents the variation of electrical conductivity versus the surfactant concentration for the m-6-6 series. The determination of the cmc is shown in the inset of Figure 1. The micelle ionization degree (R) can be estimated as the ratio of the slopes of the two straight lines above and below the cmc. These results are listed in Table 1. For comparison the data of 18-6-6 are also included in the m-6-6 series, which were measured in our previous work.30,34 Figure 2 presents the calorimetric curves of the m-6-6 series titrated into pure water, where the observed enthalpies (∆Hobs) are plotted against the final surfactant concentration. All the titration processes were found to be endothermic and show an abrupt decrease at a critical concentration. The cmc values can be determined from the intercept of the two linear extrapolations of each plot, and the enthalpies of micellization (∆Hmic) can be obtained from the difference between the observed enthalpies of the two linear segments of the plots, as shown in the inset of
11460 Langmuir, Vol. 23, No. 23, 2007
Fan et al.
Figure 1. Variation of the electrical conductivity with the surfactant concentration of the m-6-6 series at 298.15 K. The inset shows the determination of the cmc.
Figure 2. The Gibbs free energies of micellization per mole gemini molecules (∆Gmic) can be calculated from the cmc and R values.35 Furthermore, the entropy of micellization (∆Smic) can be derived from ∆Hmic and ∆Gmic. The cmc values and the thermodynamic parameters from calorimetric study are also summarized in Table 1. For comparison, the cmc values for 12-6-12 and single-chain alkyltrimethylammonium bromide (CmTAB) series are listed in Table 2. Solubilization of DMPC Vesicles by Gemini Surfactants. According to the literature,3 a schematic phase diagram of surfactant-induced lipid solubilization is illustrated in Figure 3. The total concentrations of surfactant (Dt) required for the onset (Dtsat) and completion (Dtsol) of the solubilization of phospholipid vesicles are plotted against the lipid concentration (L). Two phase boundaries separate the phases of mixed vesicles, coexisting mixed aggregates, and mixed micelles. Surfactant monomers at the onset (Dwsat) and completion (Dwsol) of the vesicle solubilization are always co-incident. Their relationship is represented by the following equations,3
Dtsat ) Dwsat + ResatL
(1)
Dtsol ) Dwsol + ResolL
(2)
In idealized form, Dtsat and Dtsol should be a linear function of L over a relatively large range of lipid concentration, and the intercepts of the two lines at Y axes Dwsat and Dwsol should have the same value. In reality, however, the linearity may be changed at lower lipid concentrations, making Dwsat smaller than Dwsol. The slopes Resat and Resol characterize the composition of the coexisting mixed vesicles and mixed micelles at the phase boundaries in terms of surfactant/lipid ratio, respectively. ITC Measurement. It is well-known that both physical and chemical processes may be accompanied by a release or uptake of heat. On this basis, the vesicle-micelle transition caused by the addition of surfactants to lipid vesicles has been first investigated by Heerklotz et al.37 using ITC, and since then many surfactant/lipid systems have been well-characterized by this method.1 In the present study, ITC was also used to study the solubilization of DMPC vesicles by the gemini surfactants. To (35) Zana, R. Langmuir 1996, 12, 1208. (36) Mosquera, V.; del Rı´o, J. M.; Attwood, D. M.; Garcı´a, M. N.; Jones, Prieto, G.; Suarez, M. J.; Sarmiento, F. J. Colloid Interface Sci. 1998, 206, 66. (37) Heerklotz, H.; Lantzsch, G.; Binder, H.; Klose, G.; Blume, A. Chem. Phys. Lett. 1995, 235, 517.
construct the phase diagrams, the experiments were conducted at different DMPC concentrations. The concentration of the surfactants in the syringe was chosen high enough to ensure the solubilization of the vesicles to be completed at the end of the titration. Figure 4 shows the solubilization curves of DMPC vesicle with various gemini surfactants at different DMPC concentrations. All of the calorimetric curves have a similar change tendency, and they all have an obvious exothermic peak. As an example, a typical raw data curve obtained from the addition of a concentrated micelle solution into DMPC vesicles is present in Figure 5A. The appearance of significant exothermic peaks in Figure 4 corresponds to the solubilization process of DMPC vesicles induced by the surfactant addition. With the initial addition of the surfactants, the surfactants are demicellized and dissolve into the vesicles. At a certain concentration, the vesicles are saturated and the further addition of the surfactants leads to the disruption of DMPC vesicles and the formation of the surfactant/DMPC mixed micelles. When all the vesicles are disrupted, further added surfactant micelles only mix with the existing mixed micelles. As shown in Figure 5B, the Dtsat and Dtsol values can be determined from the two break points at the beginning and the end of the exothermic peak, corresponding to the beginning of formation of mixed micelles and the end where all vesicles have transformed, respectively. To construct the phase boundaries, the Dtsat and Dtsol values obtained for the four gemini surfactant/DMPC systems are plotted against the DMPC concentrations in Figure 6. The straight lines are obtained from linear least-squares fitting of the experimental points. The obtained effective surfactant to lipid molar ratio (Resat and Resol) and equilibrium surfactant monomer concentration (Dwat and Dwsol) at the phase boundaries are given in Table 3. In addition, the interaction of cationic single-chain surfactant DTAB with DMPC vesicles in pure water has been studied by Majhi and Blume.10 For comparison, the Re values they obtained are also listed in Table 3. As can be seen, the Re values of the gemini surfactants are smaller than that of DTAB, while those of dissymmetric gemini series are smaller than that of a symmetric one. Among the dissymmetric series, the Re values of 16-6-6 are the lowest. Additionally, the curvature of the phase boundaries becomes obvious at low lipid concentrations for the 12-6-6/ DMPC system. This phenomenon has also been observed before for the solubilization of DMPC using bile salts in pure water18 and was thought to be due to the effect of end-caps free energy in cylinder mixed micelles with finite size, which was proposed by Lichtenberg et al.38 when considering the change in the composition of the coexisting aggregates throughout the whole coexistence range. Additionally, it can be seen from Table 3 that both Dwsat and Dwsol are much lower than the cmc values for all the gemini surfactants and do not have the same value as expected according to the basic Lichtenberg model.9 The difference between Dwsat and Dwsol increases with the increase of the corresponding cmc. This is actually the case for many solubilization processes. Lichtenberg’s model requires Dwsat ) Dwsol because the chemical potential of the surfactant monomer has to be kept constant. Keller et al.11 suggested two possible explanations for the discrepancy between Dwsat and Dwsol. One is the insufficient precision of the data, particularly in samples with low phospholipid content, and another is due to a systematic error when determining the break points on the coexistence lines from the ITC curves. (38) Roth, Y.; Opatowski, E.; Lichtenberg, D.; Kozlov, M. M. Langmuir 2000, 16, 2052.
Micellization of Cationic Gemini Surfactants
Langmuir, Vol. 23, No. 23, 2007 11461
Table 1. Critical Micelle Concentration (cmc), Micelle Ionization Degree (r), and Thermodynamic Parameters for the m-6-6 Series at 298.15 K CMC (mM) surfactants 12-6-6 14-6-6 16-6-6 18-6-6 d
conductivity 16.2 5.1 1.8 0.65 a
R
∆Hmic (kJ/mol)
∆Gmic (kJ/mol)
T∆Smic (kJ/mol)
0.39 0.28 0.31 0.29 a
-2.72 -4.92 -8.68 -14.18 b
-20.59 -29.45 c -34.91 c -41.58 c
17.9 d 24.5 d 26.2 d 27.4 d
calorimetry 16.5 4.9 1.5 0.58 b
c
a From ref. 30. b From ref. 34. c Calculated using ∆Gmic ) RT (3 - 2R) ln (2cmc) - RT ln 2; here the cmc values used are from calorimetry. Calculated from ∆Gmic ) ∆Hmic - T∆Smic.
Figure 2. Variation of the observed enthalpies (∆Hobs) with the final surfactant concentration of the m-6-6 series at 298.15 K. The inset is the determination of the cmc and ∆Hmic. Table 2. Critical Micelle Concentration (cmc) for 12-6-12 and CmTAB Series from the Literature surfactants
cmc (mM)
surfactants
cmc (mM)
12-6-12 C12TAB C14TAB
1.03 a 14.6 b 3.54 b
C16TAB C18TAB
0.89 b 0.25 c
a
From ref 21. b From ref. 36. c From ref 23.
Figure 3. Schematic phase diagram for the vesicle-micelle transition in a mixed lipid-surfactant system in idealized form.
Finally, the thermodynamic parameters for the interaction of the surfactant with DMPC are estimated from the calorimetric solubilization curves. As shown in Figure 4, the heat effect of the first several injections is endothermic for all four gemini surfactant/DMPC systems, which is caused by the demicellization of the surfactants and incorporation of the surfactants into the lipid bilayers. The endothermic effect increases or decreases slowly until a sharp change in the sign of the observed enthalpy to be exothermic. This is the region where the “extreme value” in the titration curve occurs which indicates the saturation of lipid vesicles with the surfactant molecules and the appearance of initial mixed micelles. After significantly exothermic effect, the heat effect changes to endothermic again and inclines slowly to zero. This means that the vesicle-micelle transition is already completed and the further added surfactants only experience
Figure 4. Observed enthalpy ∆Hobs versus total surfactant concentration Cs for the titration of concentrated gemini surfactant solutions into DMPC vesicles of different concentrations at 298.15 K. The DMPC concentrations (L) are indicated in the plots.
mixing with mixed micelles. Referring to the literature,10,18,39 from the ITC curves, the enthalpy change per mole of surfactant for the transfer of saturated DMPC vesicles to mixed micelles within the coexistence range ∆HT may be identified approximately as the quotient of the area under the exothermic peak between the two break points of Dtsat and Dtsol. The transfer enthalpy per micfbil can be mole of surfactant from micelles to bilayers ∆Hsurfactant (39) Heerklotz, H.; Lantzsch, G.; Binder, H.; Klose, G.; Blume, A. J. Phys. Chem. 1996, 100, 6764.
11462 Langmuir, Vol. 23, No. 23, 2007
Fan et al.
Figure 5. (A) Typical raw calorimetric curve for the titration of gemini surfactant solution into DMPC vesicles. (B) Determination of Dtsat and Dtsol values from the curves of ∆Hobs versus Cs.
approximately taken as the observed enthalpy below Dtsat, where ∆Hobs changes slowly. As an example, the transfer enthalpies obtained for the solubilization of 1 mM DMPC are shown in Table 4. As can be seen, all the values of ∆HT are negative and micfbil decrease from 16-6-6 to 12-6-6, while ∆Hsurfactant values are positive for all the systems. Turbidity Measurement. From turbidity measurement we can examine the phase transformation macroscopically and directly. The turbidity of the mixed gemini surfactant/DMPC systems was monitored by adding aliquots of a concentrated gemini micellar solution into 0.5 mM DMPC vesicle dispersion. Figure 7 shows the change of turbidity as a function of surfactant concentration. It can be seen that the initial addition of surfactants leads to a decrease in turbidity; then, after a rather flat part, the turbidity declines sharply until it reaches a constant value. The initial decrease in turbidity was also observed in the literature after addition of a very small amount of ionic surfactants to lecithin vesicles. Edwards et al.4 suggested that it was not caused by a real decrease in vesicle size. Instead, the incorporation of charged surfactants with low concentration may cause a disintegration of small clusters of aggregated vesicles which results in the reduction of turbidity. As shown in the inset of Figure 7, two critical concentrations can be determined from the turbidity curves, and it is found that these values are in accordance with the Dtsat and Dtsol values from ITC curves. Meanwhile, the significant reduction of turbidity after Dtsat is an indicator for the solubilization of mixed vesicles into mixed micelles, and the very low constant turbidity values after Dtsol indicate that the solubilization process of vesicles is ended. DLS Measurements. To obtain the aggregate size distribution at different stages, we performed DLS measurements in the systems of 0.5 mM DMPC with different surfactant concentrations. It can be seen from Figure 8 that the four gemini surfactant/ DMPC systems show similar behavior with the increase of surfactant concentration. The pure DMPC vesicles have an
Figure 6. Phase boundaries for the vesicle-micelle transitions in the gemini surfactant/DMPC mixed systems at 298.15 K. Table 3. Effective Surfactant/Lipid Molar Ratio in the Mixed Aggregates and Equilibrium Surfactant Monomer Concentration at the Phase Boundaries for the Surfactant/ DMPC Systems at 298.15 K surfactants
Resat
Resol
Dwsat (mM)
Dwsol (mM)
12-6-12 12-6-6 14-6-6 16-6-6 DTAB
0.16 0.13 0.08 0.04 0.26 a
0.33 0.28 0.20 0.10 0.36 a
0.33 5.0 1.9 0.51
0.59 8.2 2.8 1.3
a
From ref 10.
Table 4. Transfer Enthalpies Obtained from ITC Curves for the Titration of Concentrated Gemini Surfactant Solutions to 1 mM DMPC Vesicles at 298.15 K surfactant
∆HT (kJ/ mol)
micfbil (kJ/mol) ∆Hsurfactant
12-6-12 16-6-6 14-6-6 12-6-6
-5.6 -7.4 -2.7 -0.86
4.5 ( 0.3 7.2 ( 0.3 4.2 ( 0.2 1.6 ( 0.2
average hydrodynamic radius (Rh) of about 80 nm. Addition of surfactants results in a broader peak or the appearance of a small shoulder peak, which means that the polydispersity of the mixed system becomes larger than before. This may indicate that the part solubilization of vesicles results in complicated intermediate aggregates, which has been observed in the literature using cyroTEM for other surfactant/lipid systems.2 Beyond Dtsol, bimodal curves appear with a new population at Rh ) 5-10 nm. These small size distributions represent the formation of mixed lipid-
Micellization of Cationic Gemini Surfactants
Figure 7. Variations of turbidity with the addition of gemini surfactants in stepwise into 0.5 mM DMPC vesicles at 298.15 K.
Figure 8. Distributions of hydrodynamic radius (Rh) for DMPC/ gemini surfactant mixed systems at different surfactant concentrations and 0.5 mM DMPC at 298.15 K.
surfactant micelles which have weak light scattering and therefore only give rise to a small peak. It should be noted that the number of large aggregates should be very small compared to the small aggregates at higher surfactant concentration, because the amplitudes are based on scattered intensity while a large aggregate has much stronger light scattering than a small one.40 That is to say, at higher surfactant concentration the mixed micelles should be the dominant species while the proportion of large aggregates should be very low, which is also supported by the very low turbidity values.
Discussion Micellization of Dissymmetric Gemini Surfactants. The effect of dissymmetry on the cmc has been discussed in our previous paper for the gemini surfactants with the same total carbon number in the hydrophobic chains.30,34 In the present (40) Kragh-Hansen, U.; le Maire, M.; Møller, J. V. Biophys. J. 1998, 75, 2932.
Langmuir, Vol. 23, No. 23, 2007 11463
case of m-6-6 series, because one hydrophobic chain length is fixed, the increase in the length of another alkyl chain leads to an increase in dissymmetry. Therefore, the two factors could not be separated completely, and they affect the micellization process cooperatively. As shown in Tables 1 and 2, the cmc values of m-6-6 are larger than the corresponding monomeric surfactants CmTAB with the same m values, which may indicate that the “gemini effect” on the cmc is essentially eliminated. Considering that one of the alkyl chains of m-6-6 has six carbon atoms, it must be this special structure leading to its higher cmc. Previous work41 has revealed that the C6 chain is not long enough to completely insert into the hydrophobic core of micelles and part of the C6 chains will remain immersed in the water. This would lead to a reduction in the effective hydrophobicity of the C6 chains.42 What’s more, comparing with the conventional CmTAB, m-6-6 has two head groups which would cause much larger electrostatic repulsion among the molecules, and the C6 chain is not hydrophobic enough to counterwork this repulsion as a longer chain does. Both factors lead to a higher cmc for m-6-6 with respect to CmTAB. Compared to symmetric m-6-m with the same total carbon number of hydrophobic chains, the cmc of m-6-6 is lower. For example, the cmc of 18-6-6 is smaller than the cmc of 12-6-12. The results show that adding two CH2 groups to one longer chain is more efficient in lowering the cmc than adding each of them to two chains separately. Besides, it is shown in Table 1 that the micelle ionization degree R is almost constant for 14-6-6 to 18-6-6, while it is rather higher for 12-6-6. Generally, the R value relates to the surface charge density at the micelle-water interface, which derives from the packing mode of the head groups.42 In the present study, the almost constant value of R may be due to the similar packing of head groups with the similar structure and aggregation number of their micelles. However, the higher R for 12-6-6 may suggest its looser packing of head groups in the micelle surface or smaller aggregation number due to its shorter chain length. The thermodynamic parameters demonstrated in Table 1 show that ∆Hmic are all exothermic and more negative in the order of 12-6-6 to 18-6-6. In general, there are mainly four contributions to the thermodynamic functions for the gemini surfactants: the van der Waals interaction between the chains, the head group repulsion, the hydrophobic interaction, and the configuration of the spacer chain.43,44 Among these interactions, the van der Waals and hydrophobic interaction will tend to make ∆Hmic negative at the present experimental temperature of 298.15 K. It is clear that the only difference for the m-6-6 series is the increasing chain length in one of the hydrophobic chains. It is generally accepted that hydrophobic interaction is enhanced when hydrophobic chains are lengthened. For the dissymmetric gemini surfactants, the influence of dissymmetry should also be considered. The previous study for m-6-n with fixed m + n has shown that ∆Hmic becomes significantly more negative with the increase in dissymmetry m/n.34 This has been explained in terms of a large increase in the hydrophobic contributions to the micellization, which is originated from the increasingly stronger intermolecular hydrophobic interaction compared to the intramolecular interaction as m/n increases. In the present m-6-6 series, due to the lower hydrophobicity of C6 chains, the intramolecular interaction between C6 and another long side chain may be relatively weak. Therefore, intermolecular interaction (41) Bai, G. Y.; Wang, J. B.; Yan, H. K.; Li, Z. X.; Thomas, R. K. J. Phys. Chem. B 2001, 105, 3105. (42) Zana, R. J. Colloid Interface Sci. 1980, 78, 330. (43) Bai, G. Y.; Yan, H. K.; Thomas, R. K. Langmuir 2001, 17, 4501. (44) Diamant, H.; Andelman, D. Langmuir 1995, 11, 3605.
11464 Langmuir, Vol. 23, No. 23, 2007
between the longer chains of different molecules should play a dominant role. With the increase in chain length, accompanied by the increase in dissymmetry, the intermolecular hydrophobic interaction between the surfactant molecules is optimized, which would further make ∆Hmic more negative. In addition, ∆Gmic is the free energy of transferring 1 mol of surfactant molecules from the aqueous phase to the micellar pseudophase.35 More negative ∆Gmic indicates the more effective removal of the hydrophobic groups from the aqueous environment. Comparing the ∆Gmic values for m-6-6, it is clear that the micellization process is more spontaneous down the series as m increases. Vesicle to Micelle Transition. As aforementioned, the gemini surfactant-induced solubilization of DMPC vesicles has been investigated by the combination of ITC, turbidity, and DLS. The ITC measurement provided rather accurate and detailed information for the vesicle-micelle transition. From the titration curves, the phase diagrams were constructed in Figure 6. Two phase boundaries separating three pseudo-phases were found for both 12-6-12/DMPC and m-6-6/DMPC systems, indicating that they all undergo vesicle-micelle transition as the surfactant concentration is increased. Further proof for the vesicle-micelle transformation is the turbidity and DLS results. In conclusion, all the results show that the mechanism for the solubilization of DMPC vesicles induced by the gemini surfactants follows the three-stage model proposed by Lichtenberg, which is similar to the conventional surfactants. Thermodynamics of Phase Transition. The results in Table 4 show that the total transfer enthalpies ∆HT are all negative and the values decrease in the order of 16-6-6, 12-6-12, 14-6-6, and 12-6-6. This may mean that the hydrophobic interaction is the major interaction force between the DMPC lipids and surfactants. Meanwhile, ∆HT seems to associate with the total hydrophobicity of the two alkyl chains of the gemini surfactants. Furthermore, the transfer enthalpies of the surfactants from micelle to bilayer micfbil ∆Hsurfactant are all endothermic, and two possible reasons should be considered. First, to fit the lamellar packing, the cone shape of the surfactant molecule must be compensated by a lateral headgroup compression (dehydration) and lateral hydrocarbon chain expansion (fluidization), as suggested by Heerklotz et al.45 Second, the intercalation of surfactants into lipid bilayers leads to the disturbance of acyl chain packing by creating a curvature strain, causing the reduction of the order of numerous neighboring lipid molecules. The disordering effect on DMPC vesicles and the dehydration of headgroups both account for the fact that ∆ micfbil is endothermic.46 Hsurfactant Effect of Surfactant Dissymmetry on Solubilization Ability. It is well-known that lipid molecules have a cylinder shape because of the large volume of the double hydrophobic chains, whereas surfactant molecules normally look like a cone because the amphiphile headgroup requires more space than the crosssectional area of the tail. Therefore, when surfactant molecules participate into lipid bilayers, strong perturbations in packing will take place. The headgroups of surfactants interact with lipid polar fragments and their hydrophobic parts insert into the hydrophobic region of the bilayer consisting of the lipid acyl chains. On the basis of the considerations of different packing requirements for phospholipids and surfactants, Fattal et al.47 have developed a molecular model for the vesicle-micelle transition and proposed that different Re values are to be expected, depending on the chain length mismatch between phospholipids (45) Heerklotz, H.; Binder, H.; Lantzsch, G.; Klose, G.; Blume, A. J. Phys. Chem. B 1997, 101, 639. (46) Heerklotz, H.; Epand, R. M. Biophys. J. 2001, 80, 271. (47) Fattal, D. R.; Andelman, D.; Ben-shaul, A. Langmuir 1995, 11, 1154.
Fan et al.
and surfactants. The larger disparity in the hydrophobic chains favors the formation of higher curvature aggregates and also results in a lower Re value.11 Moreover, due to the hydrophobic mismatch between the lengths of the lipid acyl chains and the surfactant alkyl chain, some voids will form in the hydrophobic region when the surfactants intercalate into bilayers, which will be filled-in by trans-gauche isomerization and/or interdigitation of neighboring lipid acyl chains.31 This disturbance to ordered DMPC molecules will lead to the vesicle-micelle transition. Thus, it is generally accepted that Re values represent the ability of surfactant to solubilize lipid vesicles. According to this, the present gemini surfactants are more effective than DTAB, and the dissymmetric gemini series m-6-6 are more effective than symmetric 12-6-12. The greater effectiveness of 12-6-12 compared to DTAB for lipid solubilization may be easy understood. As stated above, Re values are defined as the critical molar ratios of surfactant to lipid in the mixed aggregates when phase transition occurs. Gemini surfactant 12-6-12 has two head groups and two hydrophobic chains, which may be taken as at least two DTAB molecules while intercalating into DMPC bilayers. Therefore, less 12-6-12 molecules would be needed relative to DTAB for pure lipid vesicles to reach the saturation or solubilization limit. The lower Re values of the m-6-6 series compared to 12-6-12 may be explained as follows. On one hand, the dissymmetry hydrophobic chains may perturb the close packing of DMPC molecules more significantly and lead to larger disparity than the symmetric one, which will create larger voids in the hydrophobic core. On the other hand, the C6 chain is not hydrophobic enough to completely insert into the hydrophobic region of aggregates, and some of them will contribute to the area of the head group region which would lead to lower critical packing parameter P ) V/al (where V and l are the volume and chain length of the hydrophobic group and a is the area of the headgroup). Normally, a surfactant with lower P value would be more efficient in solubilizing lipid vesicles. Both of the above factors would lead to instability of the lipid membranes.
Conclusions In the present work, the micellization process of a series of dissymmetric gemini surfactants m-6-6 and their interaction with DMPC vesicles have been studied. The cmc of the m-6-6 series decreases obviously with the increase in both dissymmetry and chain length. The combination of several methods has constructed the phase diagrams for the solubilization of DMPC vesicle with the gemini surfactants and revealed that the solubilization follows the three-stage model. For the solubilization of DMPC vesicles, dissymmetric m-6-6 series are more effective than symmetric 12-6-12, whereas 12-6-12 is more effective than single-chain surfactant DTAB. Among the m-6-6 series, 16-6-6 has the lowest Re values. The existence of short C6 chains and the mismatch extent of hydrophobic chains between DMPC and the surfactants may be responsible for these results. The total transfer enthalpies ∆HT during the solubilization process associate with the total hydrophobicity of the two alkyl chains of gemini surfactants. The transfer enthalpies of surfactants from micelles to bilayers micfbil ∆Hsurfactant are always endothermic due to the dehydration of headgroups and the disordering of lipid acyl chain packing during the vesicle solubilization. Acknowledgment. We are grateful for financial support from the National Natural Science Foundation of China, the National Basic Research Program, and the Royal Society (Grants. 20633010, 20573123, and 2005CB221300). LA701493S
