Interactions between 12-EOx-12 Gemini Surfactants and Pluronic ABA

Interactions between 12-EOx-12 Gemini Surfactants and Pluronic ABA Block Copolymers (F108 and P103) Studied by Isothermal Titration Calorimetry. Xingf...
0 downloads 0 Views 151KB Size
Langmuir 2004, 20, 579-586

579

Articles Interactions between 12-EOx-12 Gemini Surfactants and Pluronic ABA Block Copolymers (F108 and P103) Studied by Isothermal Titration Calorimetry Xingfu Li,† Shawn D. Wettig,‡ and Ronald E. Verrall*,† Department of Chemistry, University of Saskatchewan, 110 Science Place, Saskatoon, Saskatchewan SK S7N 5C9, Canada, and Department of Biochemistry, University of Saskatchewan, 107 Wiggins Avenue, Saskatoon, Saskatchewan SK S7N 5E5, Canada Received June 10, 2003. In Final Form: November 18, 2003 The interactions between triblock copolymers of poly(ethylene oxide) and poly(propylene oxide), P103 and F108, EOnPOmEOn, m ) 56 and n ) 17 and 132, respectively, and gemini surfactants (oligooxa)alkanediyl-R,ω-bis(dimethyldodecylammonium bromide) (12-EOx-12), x ) 0-3, have been studied in aqueous solution using isothermal titration calorimetry. The thermograms of F108 as a function of surfactant concentration show one broad peak at polymer concentrations, Cp, e0.50 wt %, below the critical micelle concentration (cmc) of the copolymer at 25 °C. It is attributed to interactions between the surfactant and the triblock copolymer monomer. The critical aggregation concentration (cac) remains constant while ∆Hmax2 and the saturation concentration, C2, increase with increasing copolymer concentration. Analysis of the cac data offers semiquantitative support that the degree of ionization of the surfactant aggregates bound to polymers is likely to be larger than that at the surfactant cmc. In P103 solutions at Cp g 0.05 wt %, two peaks appear in the thermograms and they are attributed to the interactions between the gemini surfactant and the micelle or monomeric forms of the copolymer. An origin-based nonlinear fitting program was employed to deconvolute the two peaks and to obtain estimates of peak properties. An estimate of the fraction of copolymer in aggregated form was also obtained. The enthalpy change due to interactions between the surfactants and P103 aggregates is very large compared to values obtained for traditional surfactants. This suggests that extensive reorganization of copolymer aggregates and surrounding solvent occurs during the interaction. Dehydration of the copolymers by the surfactant may also play an important step in the interaction. The endothermic enthalpy change reflecting interactions between the surfactant and polymer decreases more rapidly as the length and hydrophilic character of the spacer increases, suggesting that more favorable interactions occur with the P103 monomers having shorter PEO segments.

Introduction The physical interactions between ionic surfactants and neutral polymers in aqueous solution have been the focus of numerous experimental studies. This interest arises from the widespread use of these systems to control solution properties in applications such as cosmetics, coatings, enhanced oil recovery, and detergency. As well, some neutral amphiphilic copolymer systems can serve as models to emulate hydrophobic/hydrophilic interfaces in cellular membranes. Knowledge of the mechanism of interaction between surfactant molecules and such neutral polymers could provide important insight about the potential for use of certain surfactants as vectors in drug delivery. Water-soluble poly(ethylene oxide)-poly(propylene oxide)-poly(ethylene oxide) triblock copolymers are high molecular weight nonionic surfactants, many of which can aggregate to form micelles above their critical concentration (cmc) in aqueous solution.1-3 The formation * To whom correspondence should be addressed. Tel: +1-306966-4669. Fax: +1-306-966-4730. E-mail: [email protected]. † Department of Chemistry. ‡ Department of Biochemistry. (1) Li, Y.; Xu, R.; Bloor, D. M.; Holzwarth, J. F.; Wyn-Jones, E. Langmuir 2000, 16, 10515-10520. (2) Li, Y.; Xu, R.; Couderc, S.; Bloor, D. M.; Wyn-Jones, E.; Holzwarth, J. F. Langmuir 2001, 17, 183-188.

of block copolymer micelles is very sensitive to a temperature change. This behavior has led to the widespread use of the critical micelle temperature (cmt) as a useful property in formulations containing polymers. The temperature-dependent difference in solvation of the poly(ethylene oxide) (PEO) and poly(propylene oxide) (PPO) blocks, especially in aqueous solutions, makes these triblock copolymers particularly suitable for application in thermally reversible micellization processes. Structural studies4 show that the micelles possess a hydrophobic core consisting of the PPO blocks surrounded by an outer shell, the corona, containing hydrated PEO blocks. We have reported5 that significant changes in the self-assembly of copolymer aggregates in aqueous solution can be induced by small amounts of certain additives, in particular, the effect of the additive to displace water of hydration from the monomer or the interfacial region of the self-assembled aggregate was speculated to be a possible mechanism for such change. In a number of applications of the triblock copolymers, ionic surfactants are added to achieve colloidal stability and to control structure and rheology of the solutions. (3) Li, Y.; Xu, R.; Couderc, S.; Bloor, D. M.; Holzwarth, J. F.; WynJones, E. Langmuir 2001, 17, 5742-5747. (4) Yang, L.; Alexandridis, P.; Steytler, D. C.; Kositza, M. J.; Holzwarth, J. F. Langmuir 2000, 16, 8555-8561. (5) Wen, X. G.; Verrall, R. E.; Liu, G. J. J. Phys. Chem. B 1999, 103, 2620-2626.

10.1021/la0350204 CCC: $27.50 © 2004 American Chemical Society Published on Web 12/30/2003

580

Langmuir, Vol. 20, No. 3, 2004

The surfactants preferentially bind to the PPO block and can modify the polymer aggregation even at low concentrations. The interactions between nonionic polymers and conventional ionic surfactants in aqueous solution have been extensively studied 6-14 using a number of techniques. In contrast, there have been only a few studies of interactions between gemini surfactants and neutral polymers of PEO and PPO, triblock copolymers of poly(ethylene oxide)-poly(propylene oxide)-poly(ethylene oxide) (PEO-PPO-PEO), and hydrophobically modified polymers at low polymer concentrations.15,16 The interaction of gemini surfactants with hydrophobically modified polyacrylamides results in enthalpies of aggregation similar in magnitude to the enthalpy of micellization for the surfactant.16 As will be shown, this is in contrast to the enthalpies observed in this work which are, in some cases, an order of magnitude larger. Clearly the mechanism of interaction with the triblock copolymers is significantly different and arises from the self-aggregating behavior of the triblock copolymers as compared to the modified polyacrylamides. At high polymer concentrations the formation of polymer micelles in the presence of monomers makes the interpretation of the results of these complex systems more challenging. The present study is an attempt to extend our understanding of the interactions between gemini surfactants and the copolymer aggregates under these conditions. A new family of gemini surfactants (oligooxa)-alkanediyl-R,ω-bis(dimethyldodecylammonium bromide), C12H25(CH3)2N+(CH2(OCH2CH2)xCH2)N+(CH3)2C12H25‚ 2Br-, hereafter referred to as 12-EOx-12, having hydrophilic oligo(oxyethylene) spacer chains where x ) 0, 1, 2, and 3, was used. The objective of the study was to investigate the effect of the hydrophilic spacer on the interactions occurring between these surfactants and PEO-PPO-PEO triblock copolymers (P103 and F108) having the same length of PPO block, 56 PO units, but with varying PEO block length, 17 and 132 EO units, respectively. Isothermal titration calorimetry is a sensitive technique to characterize the energetics of these interactions. The polymer and surfactant concentrations and temperature were varied to investigate their effect on the critical aggregation concentration (cac) and the saturation concentration (C2) of the surfactant in the presence of the copolymer. A nonlinear fitting function was employed to deconvolute overlapping peaks observed in the experimental thermograms. Such data treatment makes it possible to estimate the measured enthalpy changes due to interactions between the surfactant and the micelle or the monomer components of the copolymers. Estimates of the weight fraction of micellized copolymer at higher concentrations, the cac, and C2 were obtained from this analysis. While this treatment does not provide precise values of these properties, the trends in the data obtained are meaningful and allow for a semiquantitative assess(6) Wang, G.; Olofsson, G. J. Phys. Chem. 1995, 99, 5588-5596. (7) Wang, G.; Olofsson, G. J. Phys. Chem. B 1998, 102, 9276-9283. (8) da Silva, R. C.; Olofsson, G.; Schillen, K.; Loh, W. J. Phys. Chem. B 2002, 106, 1239-1246. (9) Dai, S.; Tam, K. C.; Li, L. Macromolecules 2001, 34, 7049-7055. (10) Wang, Y. L.; Han, B. X.; Yan, H.; Cooke, D. J.; Lu, J. R.; Thomas, R. K. Langmuir 1998, 14, 6054-6058. (11) Dai, S.; Tam, K. C. J. Phys. Chem. B 2001, 105, 10759-10763. (12) Dai, S.; Tam, K. C.; Jenkins, R. D. J. Phys. Chem. B 2001, 105, 10189-10196. (13) Hecht, E.; Hoffmann, H. Langmuir 1994, 10, 86-91. (14) Hecht, E.; Mortensen, K.; Gradzielski, M.; Hoffmann, H. J. Phys. Chem. 1995, 99, 4866-4874. (15) Wettig, S. D.; Verrall, R. E. J. Colloid Interface Sci. 2001, 244, 377-385. (16) Bai, G. Y.; Wang, Y. J.; Yan, H. K.; Thomas, R. K.; Kwak, J. C. T. J. Phys. Chem. B 2002, 106, 2153-2159.

Li et al.

ment of the enthalpy profiles. The results of this work show that for copolymer systems already in an aggregated form, two distinct binding phenomena are observed: (1) a stabilization of the copolymer aggregates through a replacement of hydration water by the gemini surfactant at low surfactant concentrations and (2) a subsequent dissolution of the polymer aggregates followed by the more typical binding of surfactant to the polymer monomers (beads on a string). The stabilization of the copolymer aggregates (micelles or clusters) by conventional surfactants has not been previously reported; however, a similar mechanism whereby gemini surfactant molecules are solubilized in copolymer aggregates has been proposed.15 The results obtained in the present study suggest that the effect may be similar to the dehydration observed to occur upon the addition of inhalation anesthetics to copolymer systems.5 Experimental Section Materials and Sample Preparation. Two block copolymers of average compositions EO17PO56EO17 and EO132PO56EO132, denoted P103 and F108, respectively, were gifts from BASF and used without further purification. The nominal molar masses of these copolymers are 4640 and 16250 g mol-1, respectively. Gemini surfactants having hydrophilic oligo(oxyethylene) spacer chains, 12-EOx-12 with x ) 1-3, were synthesized as described elsewhere.17,18 The surfactants were recrystallized at least three times from ethyl acetate or acetone to give pure compounds. For comparison, 12-EO0-12 (12-2-12) was also prepared.19,20 The structures of all compounds were confirmed by 1H NMR spectroscopy and the purity was verified by surface tension measurements and elemental analysis.18,20 Stock solutions of 10 wt % of copolymers and 2-20 mM of surfactants were prepared by weight using analytical grade water obtained from a Millipore Milli-Q filtration system. Isothermal Titration Calorimetry. The titration calorimetric measurements were made using a Calorimetry Sciences Corporation model 4200 with 1.3 mL cells. The experiments consisted of a series of consecutive additions of concentrated surfactant/polymer solution to the calorimeter vessel initially containing 1.0 and 1.1 mL of polymer solution and pure water, respectively, in the sample and reference cells.21 The polymer concentration was kept constant during the experiment by using a surfactant/polymer stock solution as the titrant to prevent dilution of the polymer. However, virtually no difference in the enthalpy profiles was observed when several trials were performed in which the aqueous surfactant solution was titrated into the aqueous polymer solution. Normally, the concentrated surfactant solution was injected in either 5 or 10 µL increments into the stirred sample cell using a 250 µL Hamilton syringe controlled by the injection apparatus of the instrument. A microprocessor-controlled motor-driven syringe was used for the injections. The titration procedure was carried out using 7-min intervals between each injection. For the systems containing P103, the volume of injection was set to 3 µL at low surfactant concentrations to examine the enthalpy changes. All experiments were repeated twice and the reproducibility was within (4%.

Results and Discussion Interaction of 12-EOx-12 with F108. The F108 system was chosen for this study because it has a relatively high critical micelle concentration, cmc = 4.5 wt % at 25 °C, and provides an opportunity to observe enthalpy changes (17) Dreja, M.; Gramberg, S.; Tieke, B. Chem. Commun. 1998, 13711372. (18) Wettig, S. D.; Li, X. F.; Verrall, R. E. Langmuir 2003, 19, 36663670. (19) Zana, R.; Benrraou, M.; Rueff, R. Langmuir 1991, 7, 10721075. (20) Wettig, S. D.; Verrall, R. E. J. Colloid Interface Sci. 2001, 235, 310-316. (21) Wettig, S. D.; Nowak, P.; Verrall, R. E. Langmuir 2002, 18, 5354-5359.

Interaction of 12-EOx-12 with ABA Copolymer

Langmuir, Vol. 20, No. 3, 2004 581

Table 1. Thermodynamic Parameters for the Interaction of the 12-EOx-12 Gemini Surfactants with Pluronic F108 in Aqueous Solution surfactant 12-EOx-12 12-EO0-12 12-EO1-12 12-EO2-12

12-EO3-12 a

Cp, wt % 0.1 F108 0.1 F108 0.05 F108 0.1 F108 0.5 F108 0.5 F108c 0.1 F108

cmc,a mM

∆Hmic,b kJ/mol

0.89 1.02 1.04

-20.5 -13.5 -9.2

1.05

-9.2

∆Hmax1,b kJ/mol

Cmax1,a mM

146

0.07

cac2,a mM

∆Hmax2,b kJ/mol

Cmax2,a mM

C2,a mM

extent of binding mole ratio of Cs/Cp

0.12 0.12 0.13 0.12 0.12 0.75 0.14

31.8 28.7 19.5 26.9 55.9 11.7 25.3

0.40 0.39 0.43 0.44 0.46 1.09 0.47

2.10 2.25 1.82 2.40 2.78 2.60 2.55

32 34 55 37 8 6 39

Estimated error (0.02 mM. b Estimated error (5%. c Measured at T ) 35 °C.

Figure 1. ITC thermograms of ∼15 mM 12-EO2-12 titrated into different concentrations of F108 solutions at 25 or 35 °C. Inset shows profile for titrant concentration of ∼2 mM. The open circles are the 12-EO2-12 dilution curve in water.

arising from intermolecular interactions between the surfactants and the monomer form of the block copolymer, only. This is in contrast to the more complex system of P103, to be described in the next section, where polymer micelles exist for polymer concentrations >0.05 wt %.22-24 Plots of the enthalpy profile for the titration of ∼15 mM solutions of 12-EO2-12/copolymer into aqueous solutions of F108, 0.50 wt % or less, as a function of surfactant concentration at 25 and 35 °C are shown in Figure 1. It was convenient to use a higher surfactant concentration when titrating since it enabled one to cover the complete concentration range from the cac to C2 in a single experiment. However, the onset of the cac is not clearly shown under such conditions. The inset shows the profile for titration with a low surfactant concentration (∼2 mM) and there is clear evidence of a cac. The good agreement with the value estimated from the analysis of the experimental profile obtained by using a higher titrant concentration provides assurance that peak properties (Table 1) derived for these experimental conditions are reasonable estimates. The cac is quite low compared to results reported for conventional surfactants,1-3,11 indicating that gemini cationics behave differently from conventional cationic amphiphiles. The enthalpy profiles are endothermic at 25 °C. For comparison, the enthalpy profile of aqueous 12-EO2-12 is shown and the abrupt decrease in the profile at 1.04 mM of surfactant corresponds to its cmc. The value is in good (22) Alexandridis, P.; Holzwarth, J. F.; Hatton, T. A. Macromolecules 1994, 27, 2414-2425. (23) Ghoreishi, S. M.; Fox, G. A.; Bloor, D. M.; Holzwarth, J. F.; WynJones, E. Langmuir 1999, 15, 5474-5479. (24) Alexandridis, P.; Hatton, T. A. Colloids Surf., A 1995, 96, 1-46.

agreement with previously reported results obtained by using other techniques.17,18 The difference between the ITC curves for the 12-EO2-12/F108 and 12-EO2-12/water systems can be directly attributed to interactions between the polymer and surfactant.6,11 The surfactant concentration where this difference becomes evident, the cac, signals the onset of the formation of small 12-EO2-12 aggregates on or near the polymer. Beyond this surfactant concentration, the ∆Hobs increases sharply and reaches a maximum and then decreases with further increase in surfactant concentration, finally merging with the enthalpy curve obtained for the aqueous surfactant curve in the postmicelle region of the binary system. This behavior is typical of surfactant binding to unassociated copolymer units and is qualitatively similar to that observed in an ITC study of interactions between gemini surfactants of similar tail length, but containing spacers of three and six methylene groups, and several modified polyacryamides.16 However, it differs from the results of polymer F127/SDS studies in which the titration curves show a shallow exothermic minimum that approaches the titration curve for the aqueous surfactant system from below.2 Possible contributions to the endothermic profiles in Figure 1 (25 °C) include the dissociation of surfactant micelles from the injected concentrated surfactant solution, dilution effects, conformational changes in the polymer, dehydration of the F108 polymer monomers, and interactions between the surfactant and copolymer monomers at the conditions investigated. When one considers the difference between the enthalpy profiles of the ternary and binary aqueous systems, only the latter three of the contributions mentioned above are important. In previous studies25 of regular single-tail and single-head cationic surfactants with other nonionic copolymers, it was rationalized that beyond the cac, as more surfactant aggregates are formed and/or become larger, electrostatic repulsion between the cationic headgroups of the surfactant begins to retard aggregate growth on the polymer.12 Furthermore, it was shown that the maximum in the enthalpy curve represents the surfactant concentration at which free micelles of the surfactant begin to form in solution.25 Further addition of surfactant contributes to both free surfactant micelles and polymer-induced surfactant aggregates until the polymer becomes saturated with surfactant beads, C2. Estimates of important profile properties (Table 1) were obtained by fitting the data using an Origin based nonlinear function (Pulse), which, while not corresponding to any thermodynamic model, allows for a more accurate estimation of peak position (concentration) and magnitude (∆H). In the discussion, peaks referring to interactions between the surfactant and aggregated forms of the copolymer and between the surfactant and monomer copolymer will be labeled 1 and 2, respectively. In Table (25) Holmberg, C.; Nilsson, S.; Singh, S. K.; Sundelof, L. O. J. Phys. Chem. 1992, 96, 871-876.

582

Langmuir, Vol. 20, No. 3, 2004

Figure 2. ITC thermograms of 19.5 mM 12-EO2-12 titrated into 0.10 wt % P103 solution at 15, 25, and 35 °C.

1, the estimated values of cac2 and the surfactant concentration at the maximum, Cmax2, of the endothermic curve are seen to be somewhat insensitive to F108 concentration when the copolymer exists in monomer form. However, the magnitude of the enthalpy maximum, ∆Hmax2, and the total area of the endothermic peak, exclusive of the binary aqueous surfactant solution contribution, increase with polymer concentration. The surfactant concentration (ca. 0.45 mM) at the enthalpy maximum is somewhat lower than the cmc of regular micelles of 12-EOx-12 (0.89-1.04 mM) and reasons for this have been previously discussed.25,26 The magnitude of C2 shifts to higher surfactant concentration with increasing F108 concentration (Table 1). This is consistent with the expectation that as the polymer concentration increases there would be more hydrophobic sites for the surfactant to bind, and a greater amount of added surfactant would be required to saturate the F108 polymer chains. If one were to take the difference, C2cac2, as the amount of surfactant that interacts with the polymer in the absence of any free micelles and monomers of the surfactant, then a rough estimate of the upper bound of the mole ratio of surfactant to copolymer gives values of ca. 8 and 55 for 12-EO2-12 in 0.50 and 0.05 wt % F108 at 25 °C, respectively. However, the formation of regular micelles and the presence of monomers at surfactant concentrations beyond the peak maximum will lower these estimates. The enthalpy profiles at 25 °C do not show the presence of F108 micelles at the copolymer concentrations used since they are somewhat below the cmc. However, the enthalpy profile for 12-EO2-12 in 0.50 wt % F108 in water at 35 °C (Figure 1) is illustrative of conditions when copolymer micelles are present, since the temperature is greater than the cmt for this polymer concentration.5,22 A discussion of the changes in the enthalpy profiles due to the presence of copolymer aggregates is presented in the section, below, dealing with the effect of temperature on the interactions between gemini surfactants and block copolymer P103. Interaction of 12-EOx-12 (x ) 0, 2) with P103. 1. Effect of Temperature. The enthalpy profiles shown in Figure 2 for the interaction of 12-EO2-12 with 0.10 wt % P103 at 15, 25, and 35 °C illustrate three distinct “modes” of interaction that can be interpreted in terms of the state of P103 itself under these conditions. At 15 °C 0.10 wt % P103 exists in its monomer form, only, and the enthalpy (26) Wettig, S. D. Studies of the interaction of gemini surfactants with polymers and triblock copolymers. Ph.D. Thesis, University of Saskatchewan, 2000.

Li et al.

profile resembles that obtained for the 12-EO2-12/F108 systems (cf. Figure 1) when that copolymer is in its monomer state. The cac2 is found to be 0.27 mM, indicating that binding of the surfactant to the polymer begins at low concentrations. Upon further addition of surfactant the enthalpy isotherm reaches a maximum that corresponds to the onset of regular micelle formation and then approaches and becomes coincident, at C2, with the profile obtained for the binary surfactant system. The enthalpy profile at 25 °C shows two distinct maxima separated by a shallow minimum. Since the cmc of P103 is 0.05 wt %,24 the first peak can be attributed to interactions between the surfactant and P103 aggregates and the second with P103 monomers (similar to that observed at 15 °C). Thus the overall profile shows evidence of an initial interaction with P103 aggregates (either micelles or clusters) at very low surfactant concentrations followed by a disruption of these aggregates and subsequent interactions with P103 monomers. The present results with the gemini surfactants differ from previous results reported for systems of cetyltrimethylammonium chloride (CTAC)/P123 in which the first endothermic peak at low surfactant concentration was not completely resolved.8 One explanation for the sharp, first endothermic peak is that small additions of hydrophobic substances may induce stabilization of the PPO core by replacing water at the core-corona interface.2,5 This effect will be strong in P103 since its micelles have a smaller corona of PEO segments surrounding the hydrophobic PPO core. However, it is evident that further addition of gemini surfactant then leads to disruption of the copolymer micelles and concomitant binding of surfactant to the copolymer monomers. The breakdown of the triblock copolymer micelles by the gemini surfactant is rather more abrupt as compared to traditional single-head, single-tail surfactants. The enthalpy profile at 35 °C is somewhat different from those at 15 and 25 °C. The copolymer is in the micelle state, but the magnitude of endothermic peak 1 is considerably reduced relative to that at 25 °C. Furthermore, as the surfactant concentration increases, the enthalpy change becomes exothermic with a broad minimum at ca. 0.8 mM 12-EO2-12. With further addition of surfactant the change in enthalpy becomes endothermic, goes through a maximum, and merges with the profile for the copolymer systems at 15 and 25 °C. The shape of the enthalpy profile is similar to that observed for 12-EO2-12 in aqueous 0.50 wt % F108 at 35 °C (Figure 1). Micelles of F108 exist at this temperature since the cmt is 31.5 °C.5,22 However, the enthalpy maximum, ∆Hmax1 (Table 1), is much larger in the case of F108 than for 0.10 wt % P103. Whereas the PPO segments have the same mass in these two triblock systems, the ratio of the mass of a PEO segment in F108 to P103 is ca. 8:1. It would appear that disruption of the F108 micelles must occur with a significant increase in entropy that compensates the rather large enthalpy change associated with the overall processes reflected in this peak. This would be consistent with a major reorganization of the solvent during disruption of the copolymer micelles by the surfactant. The deeper exothermic minimum in the P103 profile compared to that of F108 at 35 °C shows that the thermal energy is more effective in dehydrating the smaller hydrophilic corona of the P103 micelles. 2. Effect of Copolymer Concentration. Figures 3 and 4 show the enthalpy profiles at 25 °C for 12-EO2-12 and 12-EO0-12, respectively, in different concentrations of P103 together with the dilution curves of the same surfactants in water, shown as open circles. The ther-

Interaction of 12-EOx-12 with ABA Copolymer

Langmuir, Vol. 20, No. 3, 2004 583

Figure 3. ITC thermograms of ∼15 mM 12-EO2-12 titrated into different concentrations of P103 solutions at 25 °C: (a) Cp ) 0.01, 0.025, 0.05, and 0.10 wt %; (b) Cp ) 0.15, 0.25, and 0.50 wt %. The open circles are the 12-EO2-12 dilution curve in water.

Figure 4. ITC thermograms of ∼15 mM 12-EO0-12 titrated into different concentrations of P103 solutions at 25 °C. The open circles are the 12-EO0-12 dilution curve in water.

mograms of 12-EO2-12 in P103 solutions of 0.01 and 0.025 wt % (Figure 3) are similar to those for F108 (cf. Figure 1), corresponding to the interaction of surfactant with the monomer polymer. At P103 concentrations in the range 0.05-0.10 wt %, the enthalpy profile shows two endothermic maxima, very clearly in the case of Cp ) 0.10 wt % but far less evident in the case of Cp ) 0.05 wt %. A similar result is shown in Figure 4 for the same P103 concentration in the presence of gemini surfactant, 12EO0-12. The two peaks become well resolved at P103 concentrations >0.10 wt %, in the presence of 12-EO2-12 (Figure 3b). On the basis of the previous discussion, peak 2 is attributed to the interaction of gemini surfactant with P103 monomer and peak 1 to the interaction of surfactant with P103 micelles or clusters and will be discussed below. Analysis of Interaction between Gemini Surfactant and P103. The utility of trying to determine the relative contributions of different components in the case of overlapping peaks in the enthalpy profile has been previously discussed.27-29 An attempt was made to obtain more information about peaks 1 and 2 through deconvolution of the overall enthalpy profile by using the Origin based nonlinear fitting function to fit the ∆Hobs data as a function of 12-EO2-12 concentration for two peaks when (27) Spink, C. H.; Chaires, J. B. J. Am. Chem. Soc. 1997, 119, 1092010928. (28) Hamm, P.; Lim, M.; DeGrado, W. F.; Hochstrasser, R. M. J. Phys. Chem. A 1999, 103, 10049-10053. (29) Li, X. F.; Imae, T.; Leisner, D.; Lopez-Quintela, M. A. J. Phys. Chem. B 2002, 106, 12170-12177.

Cp ) 0.05 and 0.10 wt %. The thermodynamic parameters for peak 2, representing the interaction between surfactant and monomeric P103 for Cp ) 0.01 and 0.025 wt %, were used to estimate the enthalpy profile of peak 2 at Cp g 0.05 wt %. This peak profile was then subtracted from the experimental enthalpy profile to obtain an estimate of the profile for peak 1, resulting from the interaction between the surfactant and the P103 micelles. The fitted lines are shown in Figures 3a and 4. The properties of peaks 1 and 2 are shown in Table 2 and discussed below. 1. Medium Concentrations of P103. Since the cmc of aqueous P103 at 25 °C is 0.05 wt %,5,24 P103 exists in its monomer form when Cp ) 0.01 and 0.025 wt % and the micelle form of the polymer is expected to dominate when Cp g 0.05 wt %. This is corroborated by the results shown in Figure 3. The presence of two peaks in the thermograms shows evidence that both monomer and micelle components of the copolymer interact with the surfactant during the course of the titration when Cp > 0.05 wt % whereas only one peak is apparent when Cp < 0.05 wt %. The enthalpy profile for Cp ) 0.10 wt % shows two overlapping components, while that for Cp ) 0.05 wt % shows evidence of a shoulder on the main peak that is consistent with there being a dynamic equilibrium between monomer and micelle species at this concentration. A further enthalpic contribution to peak 1 at copolymer concentrations close to the cmc could arise from stabilization of premicelle P103 aggregates by the surfactant (such clusters have been previously reported5). Generally, as the P103 concentration increases in the range 0.05-0.50 wt %, the magnitude of the enthalpy of peak 1 increases while the position of its maximum, Cmax1, remains approximately constant at 0.13 mM 12-EO2-12 (Table 2). On the other hand, ∆Hmax2 of peak 2 passes through a maximum value at Cp ) 0.10 wt %. At the same time Cmax2 of peak 2 remains constant at 0.35 mM 12-EO2-12 for P103 concentrations 0.025 wt %. Assuming that at Cp < 0.025 wt % there are no copolymer micelles present and for Cp ) 0.50 wt % the micelle weight fraction of polymer, Cp1/Cp, is unity, a linear plot of ∆Hmax1 vs (Cmax2 - 0.35) would offer support for this shift. A linear correlation was obtained (r ) 0.997) for 0.50 g Cp g 0.025

584

Langmuir, Vol. 20, No. 3, 2004

Li et al.

Table 2. Thermodynamic Parameters for the Interaction of the 12-EOx-12 Gemini Surfactants with Pluronic P103 in Aqueous Solution surfactant 12-EOx-12 12-EO0-12 12-EO1-12 12-EO2-12

12-EO3-12 a

Cp, wt % 0.025 P103 0.05 P103 0.1 P103 0.1 P103 0.01 P103 0.025 P103 0.05 P103 0.1 P103 0.1 P103c 0.1 P103d 0.15 P103 0.25 P103 0.5 P103 0.1 P103

cmc,a mM

∆Hmic,b kJ/mol

0.89

-20.5

1.02 1.04

1.05

-13.5 -9.9

-9.2

∆Hmax1,b kJ/mol

Cmax1,a mM

14.3 65.8 36.2

0.09 0.13 0.12

9.4 31.7

0.10 0.13

15.7 81.5 176 372 26.2

0.15 0.15 0.12 0.12 0.14

cac2,a mM

∆Hmax2,b kJ/mol

Cmax2,a mM

C2,a mM

extent of binding mole ratio of Cs/Cp

0.06 0.16 0.21 0.21 0.05 0.05 0.12 0.32 0.27 1.28 0.65 1.30 3.18 0.26

34.4 41.4 48.5 48.4 18.4 26.0 37.7 45.7 43.4 7.7 35.4 19.8 4.0 35.5

0.40 0.47 0.60 0.54 0.35 0.35 0.42 0.62 0.78 1.90 0.75 1.60 e 0.66

1.86 2.26 2.50 2.93 1.68 2.1 2.80 3.72 3.78 3.61 3.56 3.29 e 4.10

33 26 18 22 75 38 33 27 16 19 33 71 e 31

Estimated error (0.02 mM. Estimated error (5%. c Measured at T ) 15 °C. b

wt %. On the basis of this result, the weight percent of polymer in micelle form, Cp1, for copolymer concentrations g0.05 wt % were estimated by using the relation

Cp1 ) Cp1,0∆Hmax1/∆Hmax1,0

(1)

where Cp1,0 is taken to be 0.50 wt % and ∆Hmax1,0 to be 372 kJ/mol (Table 2). The weight percent of copolymer, Cp2, in the form of monomer then was estimated to be 0.04, 0.06, 0.04, and 0.01 when Cp is 0.05, 0.10, 0.15, and 0.25 wt %, respectively. Both Cp1 and Cp1/Cp increase with increasing copolymer concentration whereas Cp2, shows a maximum at Cp ) 0.10 wt %. This analysis, although based on less than stringent assumptions, does provide a rationale for the changes in peak 2 as a function of copolymer and surfactant concentrations. 2. High P103 Concentrations. The shift of the maximum of peak 2 to higher surfactant concentrations with increasing copolymer concentration reflects the fact that, once the copolymer has been transformed to its monomer form, substantially greater quantities of surfactant are then required in the formation of copolymersurfactant aggregates (peak 2). These results are consistent with the fact that the stability of aggregates of these block copolymers are very much dependent upon the delicate balance that exists between hydrophilic and hydrophobic forces in the aggregates and that this balance can be easily changed by modest temperature changes or addition of small quantities of amphiphilic materials. The very broad surfactant concentration over which peak 2 extends at the high copolymer concentrations is consistent with aggregates having low surfactant-copolymer mole ratios and the competing formation of free surfactant micelles beyond Cmax2. At the lowest surfactant copolymer mole ratio studied in this work (0.50 wt % P103), the experimental concentrations of surfactant used did not extend to sufficiently high concentrations to exhibit the presence of free micelles (cf. 0.15 and 0.25 wt % profiles in Figure 3b). On the basis of the fitting procedures described above, the extrapolated value for this parameter is estimated to be ca. 3.5 mM (Table 2). The fact that ∆Hmax2 reaches a maximum value at 0.10 wt % P103 and then decreases (less endothermic-Table 2) suggests that the surfactant:copolymer mole ratio becomes too low at higher copolymer concentrations to create substantive amounts of copolymer-surfactant aggregates (beads of surfactant on copolymer). Therefore the more energetically favorable process for the surfactant at the concentrations studied is to form free micelles, perhaps stabilized by the incorporation of hydrophobic segments of the copolymer into the core of the gemini micelle.

d

Measured at T ) 35 °C. e Not determined.

Effect of Spacer Number of 12-EOx-12. For the systems containing 0.10 wt % P103 and F108 triblock copolymer, gemini surfactants 12-EOx-12 with x ) 0-3 were introduced to examine the effect of spacer length and hydrophilic character on the interaction of surfactant with the block copolymers. The total polymer concentration was fixed at Cp ) 0.10 wt %, and the concentrations of micellized and monomeric forms of P103 were estimated (from the above discussion) to be 0.04 and 0.06 wt %, respectively. Figure 5 shows the isothermal titration curves of 12-EOx-12 in 0.10 wt % P103 and F108 solutions together with the dilution curves of the surfactants in water (open circles). The enthalpy profile for F108 at Cp ) 0.10 wt % shows a single peak and that for P103 at Cp ) 0.10 wt % shows two overlapping peaks at ca. [12-EOx12] ) 0.25 mM. For the system containing Cp ) 0.10 wt % of F108 and 12-EO0-12, the ∆Hobs increases sharply when the surfactant concentration reaches 0.12 mM (cac) and rises to a maximum at 0.40 mM and then decreases. With increasing surfactant concentration, the titration curve intersects with the 12-EO0-12/water curve, and continues to decrease to a minimum at [12-EO0-12] ) 1.42.0 mM due to the rehydration of PPO (or PEO) segments of F108 transferred from the hydrophobic core of 12-EO012 micelles to the water phase. These rehydrated segments may then wrap around the circumference of 12-EO0-12 micelles to form another type of 12-EO0-12/F108 aggregation complex resulting from ion-dipole association between 12-EO0-12 headgroups and PEO (or PPO) segments.6,7,11,12 As shown in Figure 5 and discussed previously, two overlapping peaks for the systems containing 0.10 wt % P103 can be deconvoluted from the thermogram profiles. The peak values of ∆Hmax1 and ∆Hmax2 clearly decrease with increasing spacer number from x ) 0 to 3, although the integrated areas of the thermograms are almost the same. The ∆Hmax1 for 12-EO0-12 is 66 kJ/mol and for 12EOx-12, with x ) 1, 2, and 3, is 36, 32, and 26 kJ/mol, respectively. The value of ∆Hmax1 decreases ca. 5.1 kJ/mol per EO segment in the spacer. Figure 6 shows a plot of ∆Hmax1 vs spacer number. By extrapolation, the ∆Hmax1 of 12-EO0-12 would be expected to be 41 kJ/mol, point A in Figure 6, if it followed the same interaction model. However, the experimental value is 66 kJ/mol, which is approximately twice the value for the systems containing 12-EOx-12 with x ) 1-3. This suggests that the binding between 12-EO0-12/P103 and 12-EO1-3-12/P103 are different. Peak 1 for 12-EO0-12 is the narrowest among these formed with gemini surfactants indicating that it may bind to P103 aggregates more readily and can disrupt the

Interaction of 12-EOx-12 with ABA Copolymer

Langmuir, Vol. 20, No. 3, 2004 585

Figure 5. ITC thermograms of ∼15 mM 12-EO0-3-12 titrated into 0.10 wt % P103 and F108 solutions at 25 °C: (a) 12-EO0-12, (b) 12-EO1-12, (c) 12-EO2-12, and (d) 12-EO3-12. The open circles are the corresponding 12-EOx-12 dilution curve in water.

Figure 6. ∆Hmax1 as a function of spacer number, x, in 12EOx-12 series at 25 °C.

aggregates more effectively, as compared to the other three gemini surfactants of the series that have a more hydrophilic spacer. Light scattering or cryo-TEM studies may be able to better differentiate among possible foci of interaction between the gemini surfactants and neutral copolymer. Further Thermodynamic Considerations. The Gibbs energy of surfactant-polymer interactions, ∆Gr, is usually derived from ∆Gmic and ∆Gagg (the free energy change for polymer aggregation) by assuming that the degree of micelle ionization, R, is the same in the presence of the polymer as it is in the free micelles. By using ITC enthalpy data, one can then derive (T∆S)r. However, experimental determinations of R, from EMF measurements using surfactant membrane electrodes selective to

1:1 quaternary cationic surfactants in the presence of neutral polymers,25 have shown that it varies from a rather larger value at the cac to a lower value which is approximately the average of the values at the cac and C2. The properties estimated in the present study provide an opportunity to test this assumption. Since the cac of the copolymer-gemini system is less than the cmc of the gemini micelles, formation of gemini beads on the copolymer is favored over formation of free micelles at lower surfactant concentrations. Using values of ∆Gmic calculated previously for the 12-EOx-12 aqueous binary systems,18 an estimate of the upper limit of R for these systems at the cac was obtained assuming that the lower limit for the formation of surfactant aggregates on F108 copolymer occurs when ∆Gcac equals ∆Gmic. Values of R ) 0.47 and 0.65 were obtained for the gemini surfactants with x ) 0 and 3, respectively, in aqueous 0.10 wt % F108 at 25 °C. These differ considerably from the values of 0.18 and 0.38 for the free micelles of x ) 0 and 3, respectively, in water. Furthermore, the results of unreported conductance studies of gemini surfactants 12-3-12 and 12-6-12 in aqueous F108 show that the effect of increasing copolymer concentration is to increase the value of R relative to the free micelles by as much as 50% over the concentration range 0-1.0 wt % of copolymer.26 These results offer qualitative support that R of surfactant aggregates bound to polymers is likely to be larger at the cac than at the surfactant cmc25 and any estimate of the Gibbs energy of polymer-surfactant interactions obtained by assuming otherwise should be viewed as being approximate. Conclusions At ambient temperature and in aqueous solution, block copolymer P103 aggregates to form micelles at Cp > 0.05

586

Langmuir, Vol. 20, No. 3, 2004

wt % (25 °C), while the more hydrophilic polymer, F108, exists as a monomer at Cp < 4.5 wt % (25 °C) and as micelle aggregates at Cp ) 0.50 wt % (35 °C). The addition of gemini surfactants 12-EOx-12 (x ) 0-3) to F108 solutions at 25 °C induces the formation of small surfactant aggregates on or near the dehydrated PPO segments of the F108 monomers. Titration of additional surfactant leads to the formation of regular surfactant micelles and interaction between the copolymer and the regular surfactant micelles. Both ∆Hmax2 and C2 properties of this enthalpy peak increase with increasing F108 concentration for 12-EO2-12. However, for a given F108 concentration ∆Hmax2 decreases slightly and at a diminishing rate with increasing number of EO groups in the gemini spacer whereas C2 increases at a constant rate under similar conditions. This suggests that as the spacer of the surfactant becomes more hydrophilic it is less interactive with the F108 monomers. This copolymer has very long PEO segments, and they could impede the approach of the surfactant to a hydrophobic site on the copolymer. The more hydrophilic the surfactant, the less likely it will engage in hydrophobic-like interactions with the copolymer. Analysis of the cac data offers qualitative support that the degree of ionization of the surfactant aggregates bound to polymers is likely to be larger at the cac than at the surfactant cmc.25 The addition of gemini surfactants 12-EOx-12 to P103 solutions results in the observation of one or two peaks in the enthalpy profiles due to the interaction of gemini surfactant with monomer or both monomer and micellized copolymer, respectively. At low copolymer concentrations (e0.025 wt %), the thermograms are typical of the interaction of surfactants with monomers of nonionic polymers. At higher concentrations (g0.05 wt %) there is evidence that low concentrations of surfactant may stabilize clusters or micelles by adsorbing at the corecorona interface of the P103 micelles. Subtle changes in the solvation of the triblock copolymers thus may play a more important role in the aggregation of copolymer micelles and interactions with ionic surfactants than heretofore appreciated. Further addition of surfactant produces smaller polymer micelles and unassociated copolymers.1,8,14 Small surfactant beads begin to form on or near the unassociated copolymers when the surfactant

Li et al.

concentration, Cs, is greater than Cmax1 (see Figure 3b). Qualitatively, the minimum in the titration curve between peak 1 and peak 2 can be considered to arise from the break up of the P103 micelles and the accompanying formation of surfactant beads on the unassociated copolymer.1-3,8 The broad peak 2 originates from the interaction of the surfactant with the copolymer monomers, both inherent and induced. This peak moves to higher surfactant concentrations when Cp g 0.05 wt % and can be due to an induced shift in the monomer-micelle equilibrium caused by the surfactant. This is consistent with the fact that the shape of peak 2 and the magnitude of ∆Hmax2 are observed to be similar for systems containing 0.01 or 0.25 wt % and 0.05 or 0.15 wt % P103, due to the fact these pairs of systems have similar copolymer monomer concentrations, Cp2. The peak enthalpy, ∆Hmax2, changes little upon introduction of the first EO group into the surfactant spacer. However, as additional EO segments are inserted, the rate of decrease is enhanced (Table 1). It would appear that increased length and hydrophilic character of the surfactant spacer lead to more favorable interactions with copolymer monomers that have short PEO segments. On the other hand, the ∆Hmax1 of peak 1 for 12-EO0-12 is ca. twice that for surfactants 12-EO1-3-12 implying that a gemini surfactant with a more hydrophobic spacer interacts differently with aggregated forms of P103. The very large values of the enthalpy changes observed with these systems compared to traditional surfactants16 suggest that there is extensive reorganization of copolymer aggregates and the solvent around the copolymer species during the course of these interactions. Studies using other techniques are required to confirm and explain the results obtained in this work. Acknowledgment. The authors thank the Saskatchewan Structural Sciences Centre (SSSC) for the use of the Isothermal Titration Calorimeter. Financial assistance provided by the Natural Sciences and Engineering Research Council of Canada (NSERC) is gratefully acknowledged. LA0350204