Polymer Complexes Swollen with

Sep 30, 2010 - Centre de Génie Chimique des Milieux Rhéologiquement Complexes, Université de Nancy. , ‡. Laboratoire de Chimie Physique ...
0 downloads 0 Views 2MB Size
Download PDF
13294

J. Phys. Chem. B 2010, 114, 13294–13303

Light-Responsiveness of C12E6/Polymer Complexes Swollen with Dodecane E. Rotureau,† C. Tribet,† S. Fouilloux,† P. Marchal,§ V. Sadtler,§ E. Marie-Be´gue´,‡ A. Durand,‡ and P. Perrin*,† Laboratoire de Physico-chimie des Polyme`res et des Milieux Disperse´s, UniVersite´ Pierre et Marie Curie, CNRS, ESPCI UMR 7615, 10 rue Vauquelin, F-75231 Paris, France, Laboratoire de Chimie Physique Macromole´culaire, UniVersite´ de Nancy, CNRS, ENSIC INPL, 1 rue GrandVille, F-54001 Nancy, France, and Centre de Ge´nie Chimique des Milieux Rhe´ologiquement Complexes, UniVersite´ de Nancy, CNRS, ENSIC INPL, 1 rue GrandVille, F-54001 Nancy, France ReceiVed: July 2, 2010; ReVised Manuscript ReceiVed: September 1, 2010

The association behavior of light-responsive azobenzene modified poly(sodium acrylate)s (AMPs) with C12E6 (hexa-oxyethyleneglycol n-dodecyl ether) surfactant micelles swollen with dodecane was investigated using dynamic light scattering, UV spectrophotometry, and capillary electrophoresis techniques. AMPs complexes with oligoethyleneglycol n-alkyl ether show promising properties as emulsifiers for the light-triggered control of inversion of emulsions and the present work aims at giving new insights with respect to the nature of their photoresponse. Depending on the dodecane amount, the size of the spherical surfactant micelles was varied with radii ranging from 4 to 8 nm. AMPs can be viewed as long PAANa chains bearing several randomly distributed azobenzene groups. First, the binding behavior of the AMPs chains to the micelles swollen with various amounts of oil was thoroughly studied under dark-adapted conditions, which means that most azobenzene groups are in their trans conformation (less polar than the cis conformation obtained under UV irradiation). The binding of azobenzene to surfactant micelles, which leads to the formation of AMPs/surfactant complexes, is controlled by the energy of transfer of the azobenzene moiety from water to the micelle core and by the energy of loops formation since multiple attachments of azobenzene to a single micelle are expected with long AMPs chains. We show that the change in the energy of transfer of the azobenzene group between water and micelles upon increasing the amount of dodecane within the core of micelles was quite weak (not exceeding 0.7 kT). Within the investigated range of curvature, we observed that the energy of loops formation, which decreases with increasing micelle size (decrease of curvature or increase of oil amount) was similarly weak. The effect of the presence of dodecane on the photoresponse of the complex formation was investigated. It is shown that exposure to UV light markedly weakens the association of the AMPs with surfactant within a domain of surfactant concentrations much larger for swollen micelles than for pure surfactant micelles. Consequently, we suggest that emulsion inversion triggered by light could be due to the photomodulation of the binding of AMPs to colloidal objects with various and/or specific curvatures including surfactant mesophases or small size emulsion droplets. Introduction Light-responsive polymers and surfactants are attractive command molecules for remote (light-) controlled phenomena in complex fluids. Tailored and “smart” systems have recently emerged in many applications, particularly to reversibly control properties including phase transition,1 wettability,2 gelation,3 and surface tension.4-6 Light is a nondestructive trigger, which can be quickly switched and easily focused into specific areas. At molecular scale, the photoisomerization of a chromophore group leads to significant modifications of polarity, hydrophobicity, or conformation. For instance, (E)-azobenzene groups (apolar isomer) can be reversibly converted into predominantly (Z)polar isomers.7 The self-assembly of these chromophores,8,9 or the cooperative associations and mesophase formation,10,11 are most often involved in the amplification of the system up to the macroscopic level. * Corresponding author. E-mail: [email protected]. † Universite´ Pierre et Marie Curie. § Centre de Ge´nie Chimique des Milieux Rhe´ologiquement Complexes, Universite´ de Nancy. ‡ Laboratoire de Chimie Physique Macromole´culaire, Universite´ de Nancy.

In the field of photoresponsive polymeric systems, azobenzene derivatives are widely studied as reviewed in ref 12. We have recently shown that these copolymers coupled with neutral surfactants display remarkable ability to reversibly invert emulsions upon exposure to light.13 Facile and rapid photocontrol of both stability and inversion of (macro)emulsions were achieved thanks to the tunable hydrophobic association of the polymer and surfactant molecules with respect to the oil/water interface. Optimization of the response of the emulsion system with respect to both the chemistry and formulation was reported in ref 14. In practice, however, we believe that the detailed description of emulsion stability is related not only to exchanges occurring at the droplets interface between the amphiphiles, i.e., the surfactant, the photoresponsive polymer, and the polymersurfactant complexes, but also to the presence of oil and/or water swollen mesophases (microemulsions) in the external phase of the liquid-liquid dispersion. Understanding the mechanism of the phototriggered phase inversion and stability of emulsions at molecular scale would then require accounting for the interaction of the photopolymers with interfaces of various curvatures (ranging typically from 0.5 nm-1, spherical micelles, to 0.1 µm-1, droplets) and compositions (pure surfactant film

10.1021/jp106127w  2010 American Chemical Society Published on Web 09/30/2010

Light-Responsiveness of C12E6/Polymer Complexes

J. Phys. Chem. B, Vol. 114, No. 42, 2010 13295

SCHEME 1: Structure of Azobenzene-Modified Poly(sodium acrylate)s (AMPs)a

swelling the micelles with dodecane oil. Ultimately, this work can also be viewed as a prerequisite to figure out the respective role and behavior of polymer and surfactant coemulsifiers in the stabilization of oil emulsion droplets.

a The apolar isomer (trans) is reversibly converted to the polar cis isomer by exposure to light. The photostationnary cis/trans ratio depends on the irradiation wavelength (∼ 4:1 at 365 nm; ∼1:4 at 436 nm). In dark, about 100% of the Z-isomer is obtained.

versus oil/surfactant/water film). Recent works actually suggested the importance of curvature. As a matter of fact, in micellar solutions, the dissociation of the surfactant molecules from the more polar form of the photopolymer occurs under UV exposure and association is recovered under exposure to blue light, upon switching the azobenzene chromophore to its apolar form.15,16 In contrast, at flat interfaces covered with similar surfactants, the adsorbed polymer chains were not photodissociated.15 These findings clearly demonstrate that the geometry of the surface matters quite significantly to stimulate interfaces with light. Other theoretical17,18 and experimental19,20 studies also gave evidence of the effect of surface curvature on the structure of an adsorbed polymer layer. In the present article, we report on the association of lightresponsive polymers with surfactant micelles swollen with dodecane. The nonionic surfactant, hexaethylene glycol monododecyl ether surfactant (C12E6), forms swollen micelles with dodecane oil solubilized in the micellar core. We used azobenzene-modified poly(acrylic acid)s (AMPs) as the amphiphilic polymer for which the polarity of photochrome side groups can be reversibly modified with light (as shown in Scheme 1). The polymer and surfactant associations were investigated by complementary techniques. DLS was used to determine the size of the (swollen) micelles and polymer/surfactant complexes. Capillary electrophoresis was used to measure the binding of surfactant to polymers. UV spectrophotometry provided information on the transfer of the azobenzene moieties from water to the micellar cores. Altogether, these data enable us to analyze the composition of polymer/surfactant associates according to both the size of the micelles and the presence of dodecane. The polymers bear several azobenzene groups per chain, and presumably form complexes by multipoint attachments onto the micelles. To better understand the role of the multiple binding, i.e., the formation of polymer loops, the binding of the surfactant molecules to a telomer chain bearing a single azobenzene chromophore was also investigated. In this work, we thus aim at improving our understanding of the association/dissociation mechanism of photochrome-containing polymer with spherical micellar interfaces of various compositions and curvatures, the formation of these interfaces being practically achieved by

2. Experimental Section 2.1. Materials. Boric acid and NaCl were purchased from Merck with purity g99%. The surfactant n-dodecyl hexaoxyethylene glycol monoether was supplied from Nikkol (Japan) and used as received. The synthesis and characterization of the AMPs (the structure is shown in the Scheme 1) is described in ref 16. Size exclusion chromatography measurements showed that the poly(sodium acrylate)s precursor polymers (PAANa) used in this work have number average molecular weights of 42 000 g/mol, 60 000 g/mol (Polysciences, Inc., Warrington), and 2000 g/mol (synthesized in our laboratory). Their polydispersity indexes are 2.5, 6, and 1.6, respectively. The integration level of the azobenzene chromophore groups grafted onto the PAANa (Table 1) was determined from 1H NMR (from peaks at 3 ppm and between 7 ppm and 8 ppm, i.e., R-amido CH2 and aromatic protons) and from UV-visible spectrophotometry using the extinction coefficient 2.32 × 104 (L/mol)/cm at 350 nm for the azobenzene dye.15 The AMPs are denoted m-τC6azo, with m being the number average molecular weight of the parent PAANa (in kg/mol) and τ the degree of azobenzene modification. Solutions of AMPs and surfactants in NaOH-boric acid buffer (20 mM, pH ) 9.1) and in 300 mM NaCl were stirred for 24 h in the dark prior to measurements (dark-adapted samples). 2.2. Methods. 2.2.1. Preparation of Surfactant/Dodecane Solutions. Dodecane was mixed to pure C12E6 surfactant at a ratio ranging from 25% to 120% (in mol of dodecane/mol of surfactant). The resulting liquid was added to the buffer (20 mM boric acid-NaOH, pH 9.1 with 300 mM NaCl) and then the mixture was stirred until complete dissolution. 2.2.2. Dynamic Light Scattering (DLS). Dynamic laser light scattering was carried out on an ALV/CGS-3 compact goniometer system equipped with an ALV/LSE-5003 light scattering electronic and multiple τ digital correlator and a JDS Uniphase helium-neon laser with an output power of 100 mW, which supplies vertically polarized light with a wavelength of 632.8 nm. The data were collected by monitoring the scattered light intensity at 15 °C. Temperature control within 0.1 °C was achieved using a Julabo F25 refrigerated bath circulator, and measurements were performed on solutions filtered on 0.22 µm pore size Millipore filters directly into the scattering vials. Intensities were normalized by the scattering intensity of pure decaline under same experimental conditions. Data were analyzed using the ALV-Correlator Software Version 3.0 and ALV-Fit & Plot Software provided by the manufacturer. 2.2.3. Frontal Analysis Continuous Capillary Electrophoresis. Experiments were carried out with a Beckman P/ACE system MDQ instrument equipped with a diode array multiwavelengths UVvisible detector (Beckman Instruments Fullerton, CA) operating at 15 °C and fitted with bare silica capillaries of 75 µm × 51 cm; effective length 41 cm (Chromoptic, France). The capillary was flushed daily with 0.1 M NaOH, followed by a water rinse, and finally was allowed to equilibrate with the running buffer (20 mM boric acid-NaOH, pH 9.1 with 300 mM NaCl). Pyrene, used as a micelle marker, is dissolved in the pure surfactant at a concentration of 4% (mol/ mol of surfactant). Sample solutions were prepared by mixing C12E6 stock solution (2 g/L) with a polymer stock solution (0.5 g/L) in the running buffer. The surfactant concentration range was 0.05-1.2 g/L, and the polymer concentration was 0.2 g/L.

13296

J. Phys. Chem. B, Vol. 114, No. 42, 2010

Rotureau et al.

TABLE 1: Characteristics of the Azobenzene-Modified Poly(sodium acrylate)s (AMPs) degree of azobenzene modification (mol %) obtained from 1H NMR

42-1C6azo 60-2.5C6azo 2-0.3C6azo a

molecular weight (g/mol)a

polydispersity index

protons at 3 ppm

aromatic protons

obtained from absorbance at 350 nm

42000 60000 2000

2.5 6 1.6

1.1 ( 0.1% 2.6 ( 0.2% n.d.

0.7 ( 0.1% 2.1 ( 0.2% 0.3 ( 0.1%

0.9 ( 0.1% 2.5 ( 0.2% 0.3 ( 0.1%

Number average molecular weight of the PAANa precursors.

Figure 1. Intensity weighted distribution measured at 90° of hydrodynamic radii measured by dynamic light scattering at T ) 15 °C (in 20 mM boric acid-NaOH, pH 9.1 with 300 mM NaCl)): (A) C12E6 (10 g/L) micelles swollen with dodecane. The molar amount of dodecane is quoted in the figure; (B) 42-1C6azo (0.2 g/L) without or with C12E6 at various concentrations (0.05 g/L, 0.3 g/L, 0.8 g/L). The distribution of C12E6 (5 g/L) with no polymer (also plotted in the figure) corresponds to markedly smaller objects.

Frontal analysis15,16 was carried out using the continuous electrokinetic injection mode against the running buffer, with a positive voltage (3 kV) and constant pressure (0.1 psi) applied to the inlet. The retention time of neutral species is checked by means of a reference (mesityl oxide). A first separation was run on the dark-adapted sample, which was thereafter exposed to 365 nm UV light for 20 min (1 mW/cm2) before being reanalyzed immediately after irradiation. The concentration of free C12E6 was determined from absorbance, using a calibration curve constructed with samples of known concentrations of surfactant and no polymer. 2.2.4. UV-Visible Spectrophotometry. Absorbance measurements were carried out with a UV-visible Hewlett-Packard 8453 spectrophotometer in a quartz cell. The cell was filled with a solution at 0.2 g/L AMPs in 20 mM boric acid-NaOH buffer, pH 9.1, 300 mM NaCl. Aliquots (15-100 µL) of the surfactant stock solution at 5 g/L in water were added regularly in the cell and UV-absorption spectra were recorded after each addition. To maintain a stationary fraction of 80% of trans isomer of azobenzene during the experiments, the samples were vertically and continuously irradiated under visible light (436 ( 10 nm) with a 250 W mercury arc lamp (Oriel) equipped with the appropriate interference filter. The differential absorption spectra of the samples were obtained by subtracting the spectrum of polymer with no surfactant to the spectrum in the presence of surfactant, after normalization to the dilution factors. At room temperature, solutions prepared with 0.2 g/L of 602.5C6azo in the presence of surfactant are slightly turbid,

showing that polymer/micelle complexes are poorly soluble in water. For all the experiments performed at 15 °C solutions were transparent and light scattering confirmed the absence of phase separation at low temperatures. 3. Results 3.1. Light Scattering Characterization of Swollen Micelles and Polymer/Surfactant Mixtures. Aqueous solutions of C12E6 mixed with various amounts of dodecane (cf. Methods) were characterized by dynamic light scattering (DLS). All samples were above the critical micellar concentration (CMC ∼ 0.04 g/L at 15 °C) of C12E6. The intensity weighted distributions of radii are given in the Figure 1A. The values of the radii and polydispersity index obtained using the cumulant methods are also given in Table 2. The tabulated radii values are quite close to those obtained in the Figure 1A. According to refs 21 and 22, C12E6 forms spherical micelles of hydrodynamic radius of ca. 4 nm at temperatures below 20 °C, which matches well with our measurement of samples without dodecane (mean radius ) 4.2 nm). Adding oil at 25% mol/mol (relative to surfactant) has no significant effect on the mean radius but sharpens the size distribution, as shown by the PDI, which are smaller than 0.1 in the presence of oil (Table 2). Upon increasing the amount of dodecane, the distribution slightly shifts toward larger sizes. Average radius increases with increasing the dodecane to surfactant ratio from 4.2 nm for pure micelles to ca. 7.8 nm for micelles containing 120% dodecane. Measured intermediate values for the mean radii are 5.0 and 5.5 nm for

Light-Responsiveness of C12E6/Polymer Complexes

J. Phys. Chem. B, Vol. 114, No. 42, 2010 13297

TABLE 2: Characteristics of C12E6 Micelles with and without Dodecane As Determined by Light Scatteringa % dodecane

Rh cum (nm)

PDI cum

Rh peak (nm)

0 25 50 77 120

4.9 4.3 4.8 5.5 17.4

0.252 0.064 0.051 0.045 0.483

4.4 4.2 5.0 5.5 7.8b

a

Rh cum, PDI: average radius and polydispersity index from cumulant analysis at order 2. Rh peak: average radius of the distribution from ALV’s regularization procedure shown in the Figure 1A. b Distribution shows two peaks, Figure 1A; tabulated value is the peak value of the smaller species.

TABLE 3: Dynamic Light Scattering of 0.2g/L 42-1C6azo + C12E6 + Dodecane Mixturesa C12E6 (g/L) % dodecane Rh cum (nm) PDI cum Rh peak (nm) 0 0.05 0.3 0.8 1 1 1 1

0 0 0 0 25 50 77 120

23.1 20.5 14.4 10.4 11.5 7.5 8.5 11

0.446 0.45 0.462 0.45 0.467 0.468 0.46 0.456

34.2 31.7 25.1 13.1 15.0 12.5 8.5 8.5

a Rh cum, PDI are averaged radii and polydispersity indices (3 measurements) from cumulant analysis at order 2. Rh peak is the intensity averaged Rh in the main peak (80%-93% total intensity) of the distribution determined from ALV’s regularization procedure.

amounts of dodecane of 50% and 77%, respectively. Upon increasing the oil content above 120%, a second population emerges at an average radius of 100 nm (Figure 1A insert). At a dodecane to surfactant ratio of 120%, this population of larger species contributes to about one-third of the total scattered intensity, which basically means that this population corresponds to a negligibly small number fraction of the larger particles.

At amounts of dodecane exceeding 120%, the radius of these large species drastically increases well beyond 100 nm, pointing to the formation of oil droplets rather than swollen micelles. In what follows, the binding of the micelles to the polymer chains was studied using a predominant amount of swollen micelles and, hence, a negligible amount of large droplets. Radii of ca. 30 nm -40 nm were measured using dynamic light scattering for both the polymer and polymer/surfactant mixtures (see Figure 1B, Table 3). The PDIs (also given in the table) were found to be independent of the surfactant concentration and oil content, which probably arises from the polydispersity of the polymer chains. The significant decrease of the size of the polymer coil in the presence of surfactant gives a first evidence of the association between the polymer and the surfactant. According to the cumulant radii values shown in Table 3, the dominant species in the samples are small size objects with radii as low as 8 nm, which points to the absence of significant aggregation and formation of large objects. Accordingly, the intensities scattered reflect the proportion of micelles and polymer chains. As seen in the Figure 2A,B, addition of either pure C12E6 or dodecane-containing micelles in a solution of polymer increases the intensity well above the sum of the intensities scattered by the isolated polymer and pure surfactant solutions, which also shows the formation of polymer/surfactant complexes. In addition, the intensity values also reflect the sensitivity of the complexes to exposure under UV or blue lights. For instance, 42-1C6azo/C12E6 solutions containing 0% or 50% dodecane were irradiated under UV for 10 min, and intensities were measured afterward. In this case, we observe a decrease of the intensity of around 10% as compared to the case for the same solutions equilibrated in the dark. Figure 3 also shows the light-responsiveness of polymer (421C6azo)/surfactant mixtures. The intensity variation measured at a scattering angle of 150° of successive UV and blue irradiation cycles shows that the association/dissociation process is cyclable. In the blue, most azobenzene photochromes (≈80%) are in their trans form, which optimizes the interaction of the

Figure 2. Variation of intensities (normalized by decaline) scattered by AMPs/C12E6 solutions at an angle of 150° as a function of dodecane:C12E6 ratio (in mol/mol %): (A) 42-1C6azo (0.2 g/L) and surfactant (1 g/L); (B) 60-2.5C6azo (0.2 g/L) and surfactant (1 g/L). Samples were exposed to blue light (436 nm for 20 min) or UV light (365 nm for 20 min) as quoted in the figure.

13298

J. Phys. Chem. B, Vol. 114, No. 42, 2010

Figure 3. Variation with exposure to UV and blue lights of the relative scattered intensities of 42-1C6Azo (0.2 g/L) with C12E6 at different concentrations given in the figure (angle of measurement 150°). The intensities, I, which are normalized by the intensity measured after the first exposure to blue light (I/Iblue0), were recorded after an exposure time of 20 min.

polymer chain with the micelle. The conformation switch of the photochromes to their more polar form (≈80% of the cis isomer under UV exposure) leads to a significant decrease of the affinity of the azobenzene groups for the micelles. 3.2. Degree of Transfer of the Azobenzene Groups to Micelles. The partition of the photochrome between water and micelles was measured by UV spectrophotometry. As explained in ref 16 and in the Supporting Information (SI 1, Figure S1), the spectrum of trans-azobenzene is red-shifted with decreasing the polarity of the solvent, which specifically occurs upon the transfer of the photochrome from water into the micellar cores. The variation of the absorbance (at 370 nm) in the presence of surfactant can be analyzed accordingly to yield the fraction of trans-azobenzene that has penetrated the apolar core of the micelle as a function of total surfactant concentration.16 In the following, experiments were conducted under exposure to blue light to maintain the azobenzene groups in their predominant trans form. In Figure 4A,B, the fraction of bound azobenzene groups is shown to increase with the total surfactant concentration, irrespective of the polymer structure (for the 42-1C6azo and 60-2.5C6azo polymers) and presence of dodecane. However, in the absence of dodecane, no association occurs at low surfactant concentrations, typically below the critical micelle concentration of the pure surfactant (0.04 g/L at 15 °C23). In contrast, in the presence of dodecane, some binding was detected within this range of low concentrations. Earlier onset of the association presumably reflects that micelle-like species, which actually form at lower surfactant concentrations in the presence than in the absence of dodecane are responsible for the polymer/ surfactant association. However, immediately beyond this regime of surfactant concentrations, a rapid increase of the partition of the azobenzene groups is observed with increasing the C12E6 concentration, both with dodecane and without dodecane. In this range of higher surfactant concentrations, we consider that the data, obtained in the presence or absence of dodecane, do not differ by more than uncertainty. Our analysis takes account of the slight difference observed for the curve

Rotureau et al. without dodecane in the case of the 42-1C6azo. We believe that it would be rather hazardous to comment on such small and peculiar variations. The marked slowing down of the variation of bound azobenzene with surfactant concentration unambiguously shows that a fraction of the azobenzene moieties remains unbound in the presence of excess unbound surfactant above the CMC. The azobenzene unbound fraction is higher for the 42-1C6azo (of order of 50%) than for the 60-2.5C6azo polymer (of order of 15 to 20%). The situation where hydrophobic stickers of hydrophobically modified polymers with hydrophilic backbone do not all bind to an interface/surface is not surprising, as reported in refs 15, 16, and 24-26. These results arise from a competition between the gain in energy brought by the hydrophobic unit as it adsorbs to the surface and the entropy lost caused by the confinement of hydrophilic units near the surface. More hydrophilic monomers between two successive hydrophobic stickers favor the detachment of more hydrophobic units from the surface. The polymer/ surfactant binding ratio at high surfactant concentrations thus depends significantly on the azobenzene degree of modification of the polymer, as observed in our experiments. To avoid the multiple binding of several azobenzene groups to micelles and, hence, to make the binding analysis easier, we studied the case of an AMP with low molecular weight (2-0.3C6azo). The synthesized polymer contains on average less than one azobenzene per chain, namely 0.06 azobenzene group per ∼20 monomers, i.e., 0.06 azobenzene group per average chain length. Actually, the fraction of the chains, which does not bear any photochromes, is expected to remain unbound. Among the chains able to bind to micelles, those containing a single azobenzene group are predominant. Consequently, the polymer/surfactant binding is driven by the association of only one anchor, randomly positioned in the polymer chain. As shown in the Figure 4C, which represents the fraction of bound azo groups as a function of the surfactant weight concentration, increasing the molar fraction of dodecane in micelles markedly decreases the apparent degree of binding. At a C12E6 concentration of 0.75 g/L, which is the most concentrated surfactant solution investigated here, the effect of adding dodecane to micelles is maximum. As a matter of fact, the amounts of azobenzene groups bound to swollen micelles (in the case of 50% mol/mol of dodecane for instance) and dodecane free micelles are in a ratio of about 1.6. To determine the corresponding variation of affinity constant, the data should be fitted to an isotherm model. Usually, the adsorption of small (polymer) molecules can be treated using the Langmuir isotherm. Within this framework, the lowering of bound azo upon oil addition observed in Figure 4C would correspond to a decrease of affinity on the order of 0.7 kT over the investigated range of oil amount. However, in our system, the surfactant surface is dispersed in the whole volume of the sample. Consequently, the variation of the amount of micelles caused by the addition of dodecane changes the translational entropy of the system. Under those conditions, 1:1 stoichiometry of complexes naturally accounts for this effect provided that the fraction of bound azo is plotted as a function of the molar concentration of objects. The model is, however, valid in the limit of a negligible contribution of complexes formed by multiple binding of polymer to the same object. In our experimental conditions, the azo concentration is 3.2 µM with a maximum of 30 mol % of bound azo, as shown in Figure 4C. Within the range of surfactant concentration of 0.5-0.8 g/L, the micelle concentration is on the order of 10-20 µM assuming an aggregate number Nagg of 100 in the absence of dodecane. With this 10-fold excess of total micelles to bound

Light-Responsiveness of C12E6/Polymer Complexes

J. Phys. Chem. B, Vol. 114, No. 42, 2010 13299

Figure 4. Fraction of azobenzene groups bound to micelles, obtained by spectrophotometry, in AMPs/C12E6/dodecane samples as a function of surfactant concentration: (A) 42-1C6azo; (B) 60-2.5C6azo; (C) 2-0.3C6azo with surfactant concentration in g/L; (D) 2-0.3C6azo with surfactant concentration MCO in mol/L. The dodecane amount is quoted in the figure. The polymer concentration is 0.2 g/L.

azo groups, one can reasonably expect the formation of a majority of 1:1 azo:micelle complexes. In this condition, the fraction of bound azo plotted as a function of the (swollen) micelles molar concentration (CMO) shows that the affinity constant does not vary with the amount of dodecane by more than the uncertainties (Figure 4D). The normalization of the molar concentration of micelles is detailed in the Supporting Information (SI 2). Scheme 2A gives a representation of the overall detachment of the azo groups of the 2-0.3C6azo polymer upon addition of dodecane (Figure 4C). Complexes with one azo group per micelle form (Figure 4D), and the decrease of the number of bound azo comes from the decrease of the number of micelles upon oil addition. 3.3. Polymer/Surfactant Association. Capillary electrophoresis enables us to separate free and neutral micelles from charged complexes. The UV detection of pyrene added to the

surfactant stock solution (see Experimental Section) allows for the determination of both bound and free micelle concentrations (see calibration curve in SI 3, Supporting Information). Examples of electropherograms (SI 3) show that the free surfactant concentration can be detected from the plateau height values before the zone of the polymer/surfactant complexes. The data were analyzed by assuming that the pyrene partition in micelles does not depend on the presence of the bound azobenzene. The assumption seems reasonable provided that the ratio of bound azobenzene per surfactant molecules is low. As shown below, a ratio of less than five bound azobenzene per hundred C12E6 matches well with our results (in the experimental range of excess surfactant) and with other results on similar complexes (Triton X100/C12E8) showing the predominance of surfactant molecules in the complexes.16 The ratio of bound surfactant to polymer chains is plotted as a function of the concentration of

13300

J. Phys. Chem. B, Vol. 114, No. 42, 2010

SCHEME 2: Possible Scenarios for the Effect of Dodecane on the Binding of Azobenzene Groups to Micelles in Dark-Adapted Conditionsa

a (A) In the case of single attachment (typically for the 2-0.3C6azo), the overall detachment of azo groups results from the decrease of the number of micelles upon adding dodecane. The AMP chains and micelles form azo-micelle 1:1 complexes. (B) In the case of multiple attachment (typically for the 42-1C6azo and 60-2.5C6azo), the amount of bound azobenzene groups to micelles remains unchanged upon adding dodecane.

free (i.e., unbound) surfactant for both the 42-1C6azo and 602.5C6azo polymers (Figure 5). We studied different samples with dodecane/C12E6 ratios increasing from 0 to 120 mol % of oil. For the pure surfactant with AMPs in the absence of dodecane, the binding isotherms show three successive binding regimes upon increasing the surfactant concentration: (i) “no association” at low concentrations, (ii) “cooperative binding” above a critical binding concentration of surfactant and finally, and (iii) approaching saturation. The data show that the transition between the first (i) and second (ii) association regimes occurs at surfactant concentrations that depend on the composition of the micelles. In the absence of dodecane, the surfactant concentration transition is well above the CMC of the pure surfactant (0.04 g/L at 15 °C). In the presence of dodecane, the binding occurs well below the pure surfactant CMC and the measured transition concentration does not depend on the amount of added dodecane, at least in the investigated range of dodecane/surfactant composition. Adding oil to micelles is likely to decrease the surfactant concentration required to form dodecane/surfactant mixed micelles. The azobenzene groups of the AMPs can thus bind to swollen micelles at free surfactant concentrations lower than those observed in the absence of dodecane. Although these results point to an easier binding in the presence than in the absence of dodecane, the analysis of the data in the vicinity of the CMC is rather complicated in our experimental conditions. Actually, close to the CMC, the possible demicellization of surfactants, which can induce the release of the pyrene molecules upon dilution gives an additional complexity to the quantification of the electropherograms. For that reason, the analysis presented below only focuses on the data obtained in the presence of an excess of free micelles in solution in the above-defined (iii) regime of surfactant concentrations. In this regime, the AMP chains are saturated with micelles and the bound surfactant concentrations tend toward plateau values. First, Table 4 shows that the binding of the surfactant molecules to the polymer chains is more efficient for the 60-2.5C6azo polymer than for the 42-1C6azo polymer, irrespective of both the irradiation conditions and amount of

Rotureau et al. dodecane within micelles. Second, the amplitude of the light response of the system is sensitive to the presence of dodecane. As a matter of fact, for both polymers, the values of the bound surfactant per polymer (in g/g) for dark equilibrated and UV irradiated samples are in a ratio of about 2 for amounts of dodecane within the range 25-120%. In contrast, in the absence of oil, the ratios are about 1.2 and 1.6 for the 42-1C6azo and 60-2.5C6azo polymers, respectively. The effect of light is thus more efficient in the presence than in the absence of dodecane. At this point, it is also worthwhile to give an idea of the number of azo bound per swollen micelles from the data given in Table 4. To give an upper limit of the value of bound azo per micelle, values of 1.3 and 2.2 (g of bound surfactant/g of polymer) are chosen for the 42-1C6azo and 60-2.5C6azo, respectively. In these conditions, the number of azo groups bound per micelle is ≈3.5 and ≈5.0 for the 42-1C6azo and 60-2.5C6azo, respectively. The predominance of surfactant molecules in the complexes supports the idea that the micelles should keep their spherical shape in the presence of AMP. 4. Discussion 4.1. Curvature Effect versus Affinity Change. Association of AMP onto micelles, previously described and discussed in ref 16 is a two-step process. The initial step is the attachment of one chain to a micelle by transfer of one azobenzene moiety from the polar aqueous phase into the micellar core. The association is controlled by an affinity constant, Kassoc. The affinity (Kassoc) is sensitive to the nature of hydrophobe side groups in the polymer (e.g., more or less polar), the polarity and more generally the composition of micellar core, and subtle steric interactions including repulsion between micelles and polymer backbone. The second step involves multiple anchoring of polymer via the formation of loops (as drawn in Scheme 2B, left). For flexible polymers, this multiple association, i.e., the stability of a loop, is controlled by the balance between the energy gain of azobenzene transfer (i.e., kT ln(Kassoc)) and the entropic penalty for bringing in the same core the two ends of the loop.27 Practically, this penalty depends on the average chain length between neighboring hydrophobic side groups with the result that polymers with the shortest average lengths bind more tightly. In the present work, both AMPs with the long chains contain several azobenzene groups and can accordingly form loops. As expected, the higher fraction of azobenzene bound to micelles (Figure 4) was found for the 60-2.5C6azo polymer, i.e., the AMP with the shorter distance between azobenzene moieties as compared to the 42-1C6azo polymer. The polymer with short chain length (2-0.3C6azo) contains about 20 monomers per chain (cf. Supporting Information) and has on average 0.5 azobenzene per chain. Consequently, isolated chains contain at most one azobenzene and cannot form loops. The binding of 2-0.3C6azo is thus expected to be essentially driven by Kassoc. Two major parameters were varied simultaneously in this experiment: the curvature of the micelles and the composition of micelles core. It is a priori difficult to predict the effect of the micelle composition on the AMP/surfactant binding. In contrast, the energy of association of polymers on colloid particles with various sizes is documented. It is known that polymer chains are forced to adapt their conformation to the shape and size of the host particle to adsorb to an interface.17,28,29 Scaling approaches predict a better stabilization of the complexes with decreasing radius of particles. Compared to larger particles, smaller ones allow partial relaxation of the osmotic energy arising from the repulsion between the polymer segments of the particle corona. Although the complete calculation has only

Light-Responsiveness of C12E6/Polymer Complexes

Figure 5

J. Phys. Chem. B, Vol. 114, No. 42, 2010 13301

13302

J. Phys. Chem. B, Vol. 114, No. 42, 2010

Rotureau et al.

Figure 5. Binding isotherms of C12E6 to AMPs (0.2 g/L) in boric acid-NaOH 20 mM pH 9.1, NaCl 300 mM, as obtained by electrophoresis: (() dark adapted samples; (2) samples incubated for 20 min under exposure to UV light; (A) 42-1C6azo and no dodecane; (B) 42-1C6azo and dodecane/ C12E6 equal to 50%; (C) 42-1C6azo dodecane/C12E6 equal to 120%; (D) 60-2.5C6azo, no dodecane; (E) 60-2.5C6azo and dodecane/C12E6 equal to 25%; (F) 60-2.5C6azo and dodecane/C12E6 equal to 50%; (G) 60-2.5C6azo and dodecane/C12E6 equal to 120%.

TABLE 4: Ratio of Bound Surfactant to Polymer in the Presence of Excess of Micellesa

42-1C6azo 60-2.5C6azo

amount of dodecane (mol %)

dark equilibrated samples bound surfactant/polymer (g/g)

UV irradiated samples bound surfactant/polymer (g/g)

0 50 120 0 25 50 120

1.7 1.3 1.6 2.8 2.2 2.3 2.5

1.4 0.6 0.7 1.7 1.2 1.2 1.0

a Values are the average of measurements for surfactant concentrations in the 0.5-0.8 g/L range. The cis isomer content under UV is ≈80%.

been performed on neutral chains, the overall trend predicted by the theory should be valid for our polyelectrolyte chains since the osmotic relaxation including Coulombic terms favors the binding to smaller particles. Eventually, the effect of osmotic relaxation hampers the binding to larger particles, which in our case would correspond to a decreasing apparent affinity with increasing amount of dodecane. The [Kassoc(no dodecane)/Kassoc(with dodecane)] ratio can be determined as the ratio of the slopes of the curves given in the Figure 4D at low surfactant concentrations. Results show that the affinity constant does not change upon addition of dodecane. Assuming that the above analysis remains valid for longer chains, it becomes now possible to analyze the effect of the curvature occurring on segments of chains forming loops (polymers with multiple attachments to the same micelle). Since the isotherm curves actually display no variation upon changing the dodecane amount as the micellar radius increases from 4 to 8 nm (Figure 4A,B), we come to the conclusion that the increase of the curvature, which favors loop formation, is also not significant. The same conclusion can be drawn from the plateau values of surfactant binding isotherms in the dark (Table 4, Figure 5A-C), which do not show variations by more than

uncertainty upon changing dodecane amount. The scenario is sketched in Scheme 2B. Theories dealing with the effect of curvature on the equilibrium conformation of polymers at interface most typically span a large range of curvatures, from spherical to planar geometries.17,28,29 The amount of adsorbed chains depends on the curvature as the size of the particles varies from size smaller to larger than the polymer dimension. Since our experimental data cover regimes for which the micelle radii (e8 nm) is always smaller than the AMPs coil size (R ∼ 20-30 nm), weak effects of curvature are predicted by the theories. Our data show that this is indeed the case. 4.2. Photoresponse in the Presence of Dodecane. Photomodulation of the complexes comes from variation of polarity of the azobenzene moieties between their trans and cis forms, which in turn affects the energy of transfer from water to micelles. Consequently, the overall binding strength between AMPs and surfactants is a key parameter to achieve large responses to light. As a matter of fact, a too strong association is likely to prevent the photodetachment. From this point of view, optimization and achievement of the response to light are indeed favored by a slight decrease of the energy of transfer. Under such conditions, improvement of the photoresponse in the presence of dodecane betrays the fact that the energy of transfer of the azobenzene cis isomer should be more sensitive to the presence of dodecane and decreases more rapidly than that of the trans isomer. Table 4 and Figure 5 show that the dissociation of the AMPs/surfactant complexes with an excess of micelles is actually more pronounced in the presence of dodecane, irrespective of the modification degree of the polymer. Moreover, the presence of dodecane within the micelles also extends the window of compositions with respect to surfactant concentrations, for which marked photoresponses are achieved. This observation might be a step toward the understanding of the emulsions photoinversion mechanism.13,14 As a matter of fact, it was shown that the surfactant concentration is a limitation toward the reversible control of the formation of AMPs/ surfactant complexes (without dodecane) by switching light

Light-Responsiveness of C12E6/Polymer Complexes wavelength (blue T UV) (Figure 5A,D). On the one hand, the photocontrol of such complexes can only be achieved at surfactant concentrations much lower than those required to destabilize or invert emulsions.13,14 On the other hand, the photodetachment of AMPs polymers from flat surfactant layers was not observed, as reported previously.15 As discussed above, the presence of only a small amount of dodecane allows photocontrol of microemulsions over a broader range of surfactant compositions. Consequently, since emulsion photoinversion occurs at temperatures for which a broad range of mesophases are likely to coexist, we suggest that the photoactivation of the corresponding AMPs/mesophases could be at the origin of the light-induced emulsion inversion. In the future, it would hence be interesting to study the effect of light on the binding of AMPs to mesophases of various curvatures (typically from 10-1 nm-1 to 0), which is out of the scope of this contribution. 5. Conclusion We reported on the binding mechanism of AMPs (azobenzene modified PAANa) to spherical C12E6 micelles. The formation of the polymer/surfactant complexes was studied as a function of light (UV-blue or in the dark), composition (dodecane oil swollen and not swollen micelles), and curvature (radii in the 4-8 nm range) of surfactant micelles using DLS, UV spectrophotometry, and capillary electrophoresis complementary techniques. The binding of AMPs to the surfactant micellar aggregates is a two-step process, which depends on both the affinity of the photochrome for the surfactant-covered interface layer and formation of polymer loops. The latter effect has to be considered because our photoresponsive AMPs with high molecular weights are modified PAANa with several azobenzene groups per chain (42-1C6azo and 60-2.5C6azo). For both 421C6azo and 60-2.5C6azo, we observed a lack of sensitivity of the binding isotherm to the micellar composition. With the 2-0.3C6azo polymer, we showed that the decrease of affinity is very weak. Interestingly, it thus appears from this study that curvature (in the range 0.25-0.12 nm-1) does not significantly change the binding of both chain ends and segments of AMPs to surfactant micelles. An attempt is currently made in our group to extend this work to the binding of AMPs to objects with smaller curvatures such as oil swollen surfactant mesophases and/or emulsion droplets. The effect of dodecane on the photoresponse of the [AMPs (42-1C6azo and 60-2.5C6azo)]/[C12E6 swollen micelles] complex formation was also investigated. The photodetachment of the AMPs/surfactant complexes, which requires an optimized strength with respect to the polymer/surfactant association, were found to be easier in the presence than in the absence of dodecane within the core of micelles. Also, our data show that the surfactant domain of composition for which the AMPs/ surfactant systems exhibit marked responses to light is actually more important in the presence than in the absence of dodecane. As a matter of fact, previous studies showed that responses to light using the AMPs/pure surfactant (i.e., without dodecane) could only be achieved at rather low surfactant concentrations. In particular, the surfactant concentrations were much lower than those required to reversibly invert emulsions by switching the light wavelength.16 Furthermore, since the photoadsorption/ desorption of AMPs from flat surfactant adsorbed monolayer could not be achieved either, explaining the light-induced control of emulsion type by the direct photostimulation of emulsion droplets would also appear rather controversial.15 However, since

J. Phys. Chem. B, Vol. 114, No. 42, 2010 13303 the photomodulation of the AMPs/dodecane swollen micelles is much less restrictive with respect to formulation composition, we suggest that the reversible control of emulsion polarity with light at constant temperature could be due to molecular mechanisms involving the presence of swollen surfactant mesophases or small size emulsion droplets with various and/ or specific curvatures. As already mentioned, it would thus be of great interest to explore the binding behavior of AMPs to colloids with curvatures, which typically cover the 10 nm-1 (swollen micelles) to ≈0 (macroemulsion droplets ≈ completely flat interfaces) broad range. Acknowledgment. We thank the BLAN-07-0278 and BLAN06-0174 ANR (National Research Agency) programs. Supporting Information Available: SI 1 details the determination by UV-vis spectrometry of the % transfer of azobenzene from water into hydrophobic environment. SI 2 gives the calculations for the determination of the molar concentration of micelles to plot the Figure 4D from Figure 4C assuming that the packing of the surfactant in the free micelle and in swollen micelles remains the same. SI 3 shows the calibration curve of capillary electrophoresis and typical examples of electropherograms. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Jochum, F. D.; Theato, P. Polymer 2009, 50, 3079. (2) Lim, H. S.; Han, J. T.; Kwak, D.; Jin, M.; Cho, K. J. Am. Chem. Soc. 2006, 128, 14458. (3) Irie, M. Macromolecules 1986, 19, 2890. (4) Rosslee, C.; Abbott, N. L. Curr. Opin. Colloid Interface Sci. 2000, 5, 81. (5) Abbott, S.; Ralston, J.; Reynolds, G.; Hayes, R. Langmuir 1999, 15, 8923. (6) Yuan, W.; Jiang, G.; Wang, J.; Wang, G.; Song, Y.; Jiang, L. Macromolecules 2006, 39, 1300. (7) Porcar, I.; Tribet, C.; Perrin, P. Langmuir 2001, 17, 6905. (8) Wang, G.; Tong, X.; Zhao, Y. Macromolecules 2004, 37, 8911. (9) Zhao, Y. Chem. Record 2007, 7, 286. (10) Eastoe, J.; Vesperinas, A. Soft Matter 2005, 1, 338. (11) Eastoe, J.; Wyatt, P.; Sanchez-Dominguez, M.; Vesperinas, A.; Paul, A.; Heenan, R. K.; Grillo, I. Chem. Commun. 2005, 2785. (12) Ercole, F.; Davis, T. P.; Evans, R. A. Polym. Chem. 2010, 1, 37. (13) Khoukh, S.; Perrin, P.; Bes de Berc, F.; Tribet, C. ChemPhysChem 2005, 6, 2009. (14) Khoukh, S.; Tribet, C.; Perrin, P. Colloid Surf. A-Physicochem. Eng. Asp. 2006, 288, 121. (15) Khoukh, S.; Oda, R.; Labrot, T.; Perrin, P.; Tribet, C. Langmuir 2007, 23, 94. (16) Ruchmann, J.; Fouilloux, S.; Tribet, C. Soft Matter 2008, 4, 2098. (17) Aubouy, M.; Raphael, E. Macromolecules 1998, 31, 4357. (18) Fleer, G. J.; Cohen Stuart, M. A.; Scheutjens, J. M. H. M.; Cosgrove, T.; Vincent, B. Polymers at Interfaces; Chapman and Hall: London, 1993. (19) Greenwood, R.; Luckham, P. F.; Gregory, T. Colloids Surf., A 1995, 98, 117. (20) Cosgrove, T.; Griffiths, P. C.; Lloyd, P. M. Langmuir 2002, 11, 1457. (21) Brown, W.; Rymden, R. J. Phys. Chem. 1987, 91, 3565. (22) Mitchell, D. J.; Tiddy, G. J. T.; Waring, L.; Bostock, T.; Mc Donald, M. P. J. Chem. Soc., Faraday Trans 1983, 79, 975. (23) Goto, A.; Takemoto, M.; Endo, F. Bull. Chem. Soc. Jpn. 1985, 58, 247. (24) Millet, F.; Perrin, P.; Benattar, J. J. Phys. ReV. E 1999, 60, 2045. (25) Millet, F.; Benattar, J. J.; Perrin, P. Macromolecules 2001, 34, 7076. (26) Millet, F.; Perrin, P.; Merlange, M.; Benattar, J. J. Langmuir 2002, 18, 8824. (27) Lairez, D.; Adam, M.; Carton, J. P.; Raspaud, E. Macromolecules 1997, 30, 6798. (28) Aubouy, M. Phys. ReV. E 1997, 56, 3370. (29) Guiselin, O. Europhys. Lett. 1992, 17, 225.

JP106127W