The surfactant sodium bis(2-ethyl-hexyl) sulfosuccinate, named Aerosol OT or AOT, was from Fluka. We use the purification method published elsewhere [35] and the surfactant purity was evaluated using gas chromatography [36]. The polymer poly(maleic anhydride-alt-1-octadecen), PMAO (M_r = 40 kDa) was from Sigma-Aldrich and was used as received without further purification. The aqueous solution of poly(allylamine) 20% w:w, PA (M_r = 17 kDa) was from Sigma-Aldrich. The polymer molecular weights were provided by the manufacturer.

To obtain spontaneous aggregates, the surfactant solutions were prepared without external energy input. The AOT/PA vesicles were prepared by dissolving the surfactant in the aqueous PA solutions with a given polymer concentration. We use two different procedures for systems containing AOT and PMAO because PMAO is a water-insoluble molecule. In the first one, the aggregates were prepared by weight both, the surfactant and polymer molecules, and dissolving them in water. In the second procedure the polymer PMAO was dissolved in surfactant solutions with a given concentration. DLS measurements show that the size of the aggregates prepared by the two different methods agrees with one another. The solutions were prepared with water purified using the combination of RiOs and Milli-Q systems from Millipore.

To study the effect of polymer concentration on the properties of Aerosol OT vesicles, the aggregates were prepared with 0.03 M of Aerosol OT and different polymer concentration. The maximum polymer concentrations used were 0.020% for PA and 0.094% for PMAO, because the systems become instable above these polymer concentrations. In the case of PMAO/Aerosol OT mixtures, above 0.1% of PMAO the polymer molecules precipitate, probably because vesicles cannot incorporate more PMAO inside their bilayers. On the other hand, large flocs are clearly observed in vesicles prepared with (PA) > 0.02%, the non-uniform adsorption of PA on the vesicle surface can interpret this behavior, see below. All solutions were prepared the day before to obtain experimental measurements and were maintained at 30 °C. The measurements were carried out with clear solutions.

The electrical conductivity was measured with a conductometer model 727 from Metrohm operated at 2.4 kHz. A Metrohm Herisau conductivity cell, model AG 9101, was used. The cell constant, 0.847 cm⁻¹, was obtained by calibration with KCl solutions of known concentrations [37]. Because the correct determination of the cvc requires a large number of experimental data, a conductometric titration was employed. During titration, solutions obtained by successive dilutions were allowed to equilibrate a few minutes until a stable measurement was obtained [38,39].

Dynamic light scattering was performed with a CGS-8 from ALV GmbH working in pseudo cross-correlation mode, using the green line (λ = 514.5 nm) of an argon ion laser (Coherent I300). The intensity correlation functions were obtained at a fixed temperature of 30.0 °C and as a function of thescattering angle, θ, between 30 and 140°, corresponding to wavevectors, q, from 8.42 × 10⁴ to 3.14 × 10 5 cm⁻¹ defined as: q = 4 π n λ sin ( θ 2 ); where n represents the solution refractive index.

The normalized second order correlation functions (g⁽²⁾(t)) were analyzed using both GENDIST [40] and CONTIN [41,42] inverse Laplace algorithms, giving both similar relaxation time distributions. The solution chosen was the most probable one and corresponds with a value of the regularization parameter, α= 0.5. From the average relaxation times, τ, the apparent diffusion coefficients, D^app, were obtained and using the Stokes-Einstein relation, the apparent hydrodynamic radii were calculated [43].

The electrophoretic mobility measurements were carried out by means of the laser Doppler electrophoresis technique using the Zetasizer 3000 device (Malvern, UK). All experiments were made in a 5 mm × 2 mm rectangular quartz capillary. Each experimental value is the average of ten measurements and the standard deviation of these measurements was considered the experimentalerror. The electrophoretic mobility values were also obtained at 30.0 °C. The electrophoretic mobility, μ_e, was converted into zeta-potential, ζ, using the Smoluchowski's relation, ζ = μ e η ɛ where η and ɛ are the viscosity and the permittivity of the solvent, respectively.

In previous works we used the 4-[6-methoxy-2-napthyl]-2-butanone, nabumetone, as fluorescent probe to obtain the critical vesicle concentrations of Aerosol OT vesicles [9,29]. Even though this probe has been successfully used to study different kinds of surfactant aggregation processes [9,29,44,45], it was not used in our systems because the PA and PMAO polymers present fluorescence emission. Therefore, we use electrical conductivity measurements to obtain the critical vesicle concentration.

Figure 1 shows the experimental conductivity values as a function of the Aerosol OT for aqueous solutions of mixed polymer-Aerosol OT with polymer concentration constant. The vesicle critical concentration (cvc) is assigned with the break point of the experimental data [8]. For the sake of clarity, the figure presents results for some polymer-surfactant systems, a similar trend was observed for the rest of mixtures.

The results show two different trends as a function of the polymer added to the Aerosol OT vesicles. Thus, the cvc values of mixed PMAO-Aerosol OT vesicles are independent of polymer concentration and the average value found is (7.0 ± 0.5) mM. This value is quite similar to the cvc of pure Aerosol OT vesicles [8]. In contrast, the cvc values obtained for PA-Aerosol OT mixed vesicles are lower than the value found for pure Aerosol OT and they are almost independent of the polymer concentration. The cvc average value calculated was (2.7 ± 0.5) mM. This is the classical behavior of strongly interacting polymer-surfactant mixtures [46]. It is also interesting to note that the cvc of PA-Aerosol OT mixed vesicles is in the order of magnitude of the cvc obtained previously for stable mixed vesicles of PEG/AOT or PSS/AOT [8,9].

From the results it is possible to conclude that the inclusion of PMAO on the vesicle bilayer does not affect the critical vesicle concentration of Aerosol OT, while a significant lowering of the cvc is observed when vesicles are prepared with oppositely charged polymer and surfactant molecules. This means that the surfactant Aerosol OT molecules prefer to aggregate with the polymer PA than themselves [46].

In order to obtain information about the effect of polymer addition on the aggregates size, dynamic light-scattering experiments have been performed at 30.0 °C with different polymer concentrations. The Aerosol OT concentration was kept constant at 0.03M. The hydrodynamic radius of pure Aerosol OT vesicle, R H app = ( 69 ± 2 ) nm, was previously determined using dynamic light scattering measurements [29] and it is independent of the surfactant concentration for [AOT] ≤ 0.03 M.

The correlation functions for AOT/PMAO systems are monomodal and the inverse time decay shows a linear dependence with q², Figure 2(a). The apparent hydrodynamic radius was calculated from the apparent diffusion coefficient obtained from the slope of inverse time decay vs. q² and the Stokes-Einstein equation. The apparent hydrodynamic radius is represented against the polymer concentration in the Figure 2(b). Examination of results in the Figure 2(b) shows that the hydrodynamic radius is almost independent of polymer concentration at low polymer concentrations with a size quite similar to the Aerosol OT vesicles. This fact seems to indicate that the incorporation of small concentrations of PMAO inside the vesicles bilayer does not affect the size of Aerosol OT vesicles. However, when the polymer concentration incorporates onto the bilayer is further increased the apparent hydrodynamic radius increases from 70 to 118 nm.

All the experiments corresponding to mixtures of Aerosol OT with PA present single exponential correlation functions that, when analyzed using regularized Inverse Laplace Transforms (ILT) (GENDIST, CONTIN), lead to single peak decay time distributions. The inverse time decay shows the expected linear dependence with q². From the linear dependence of the inverse decay time with q² we calculate the apparent diffusion coefficient and using the Stokes-Einstein relation we found the apparent hydrodynamic radius. The R H app values are represented against PA concentration in the Figure 3.

Examination of Figure 3 shows that for relatively small polymer concentrations, [PA] < 1 × 10⁻⁴% w:w, vesicles have the same size than the Aerosol OT ones and when PA concentration increases up to this value, the hydrodynamic radius decreases with PA concentration until it reaches a constant value of (50 ± 1) nm at the polymer concentration of 4 × 10⁻⁴%.

It was not possible to obtain information about the morphology of the different aggregates by means of Cryo-TEM, because no clear images were obtained in these systems. This fact was previously observed in mixtures of Aerosol OT with PEG polymers [29] and it was attributed to association processes between vesicles in the freezing procedure [47,48].

The effect of the polymer addition on the zeta potential of Aerosol OT vesicles was also investigated. Figure 4 presents the ζ-potential values of vesicles prepared with 0.03 M of Aerosol OT and different PMAO and PA concentrations. For comparison, Figure 4 also presents the ζ-potential of 0.03 M Aerosol OT vesicles, shown by the open circle in the figure.

Results in Figure 4 show that in mixtures containing low PMAO concentrations, the ζ-potential remains constant in the value of pure Aerosol OT vesicles. This fact indicates that in this polymer concentration range, mixed vesicles are quite similar to the pure Aerosol OT ones. When the PMAO concentration is further increased the ζ-potential weakly decreases. This fact agrees very well with results obtained by other authors [49,50]. In these works, the combination of light scattering and ζ-potential measurements indicated that the zeta-potential often decreases with increasing the vesicle radius.

The ζ-potential values for PA/Aerosol OT vesicles are almost independent of polymer concentration until the polymer concentration reaches a value of 0.005% w:w. Above this polymer concentration, the ζ-potential of vesicles sharply decreases indicating that the polycation PA partially neutralizes the surface negative charge of the Aerosol OT vesicles.

It was not possible to obtain the isoelectric conditions, because when the polymer concentration reaches a value above 0.02% large flocs are observed. It is interesting to note that, at this polymer concentration, the vesicle electric charge is not completely neutralized. This phenomenon cannot be interpreted by the framework of DLVO theory, because in this theory the particle surface is assumed to be uniformly charged [51]; however, this assumption could not be fulfilled in the case of polyelectrolyte adsorption onto the surface of particles. It is well established that the adsorption of polyelectrolytes on the oppositely charged surface takes place in a highly correlated manner resulting in a non-uniform distribution of the adsorbed chains and, consequently, of the surface electric charge. Volegol and Twar recently developed a model for the potential of mean force between non-uniformed charged particles. They showed that, the inter-particle potential has an attractive component even in particles bearing an electric charge of the same sign. This potential is responsible of the evolution from the initial monodisperse vesicles toward a broader distribution of large aggregates that finally flocculate [52]. According to this model the relation between the ζ-potential and the radius of stable aggregates (R) is given by [53]: (1) R = 10 k B T π ɛ ln 4 ( ζ ) 2where k_B represents the Boltzmann's constant and ɛ is the permittivity of water. Taking into account that the radius found in this work for stable Aerosol OT/PA mixed vesicles is 50 nm, we use the Equation (1) to estimate the ζ-potential of vesicles with this radius. The ζ-potential value estimated from Equation (1) is ± 17 mV. According to the Volegol's model vesicles with ζ-potential values above | 17 | mV are stable. Our results are consistent with this model because the ζ-potential values measured for Aerosol OT/PA vesicles prior flocculation were higher, in absolute value, than 17 mV. On the other hand, from results in the Figure 4 we can estimate the polymer concentration needed to reach the ζ-potential corresponding to unstable Aerosol OT/PA vesicles (−17 mV). The estimated value was 0.025%, in good agreement with the phase separation observed for PA concentrations above 0.02%.