5478
J. Phys. Chem. B 2009, 113, 5478–5486
Structural Characterization of Nonionic Mixed Micelles Formed by C12EO6 Surfactant and P123 Triblock Copolymer David Lo¨f,†,‡ Matija Tomsˇicˇ,| Otto Glatter,§ Gerhard Fritz-Popovski,§ and Karin Schille´n*,† DiVision of Physical Chemistry, Center for Chemistry and Chemical Engineering, P.O. Box 124, Lund UniVersity, SE-22100 Lund, Sweden, Faculty of Chemistry and Chemical Technology, UniVersity of Ljubljana, AsˇkercˇeVa 5, SI-1001 Ljubljana, SloVenia, and Institute of Chemistry, UniVersity of Graz, Heinrichstrasse 28, A-8010 Graz, Austria ReceiVed: September 23, 2008; ReVised Manuscript ReceiVed: February 4, 2009
A structural characterization of mixed micelles formed in aqueous solution by the PEO-PPO-PEO triblock copolymer P123 and the nonionic surfactant C12EO6 was carried out using various techniques, including ultralow shear viscosimetry, depolarized dynamic light scattering (VH-DLS), depolarized static light scattering (VH-SLS), and small-angle X-ray scattering (SAXS). The sphere-to-rod transition of the mixed micelles was studied in a diluted regime (P123 concentrations ranging from 0.5 to 10 wt %) at C12EO6/P123 molar ratios of 2.2, 3.2, 6.0, and 11 as well as for the pure C12EO6. The data from VH-SLS and viscosimetry displayed a sharp increase in the intensity and viscosity, respectively, at the sphere-to-rod transition, and the results from the two methods were in accordance. In both techniques, an increased transition temperature with increasing content of C12EO6 (in the molar ratio regime from 2.2 to 11) was observed. SAXS was used as the main technique, and a thorough structural characterization was performed, where indirect Fourier transformation (IFT) and generalized indirect Fourier transformation (GIFT) were employed in the analysis procedure of the SAXS data. The p(r) functions obtained from the IFT (employed at low P123 concentrations, i.e., 1.0 and 2.0 wt %) and GIFT (employed above 2.0 wt %) analyses revealed increased inhomogeneities in the mixed micelles when the molar ratio was increased. This suggested that the C12EO6 organized themselves at the interface between the PPO core and the PEO corona of the P123 micelles, with the C12 alkyl chain stretching into the hydrophobic core and the EO6 part residing in the hydrophilic corona. The structure factor parameters obtained with GIFT for a molar ratio of 2.2 at a P123 concentration of 5.0 wt % showed radius values smaller than what was estimated from the p(r) functions. This was explained by an interpenetration of the PEO chains from one mixed micelle into a neighboring one. VH-DLS was performed on the mixed micelles at a temperature 3 °C above the transition temperature and at a molar ratio of 2.2. From the analyzed data, the average length L of the rods was estimated to be 102 nm. Introduction The self-assembly in water of nonionic amphiphiles such as triblock copolymers of poly(ethylene oxide) (PEO) and poly(propylene oxide) (PPO) in the form PEO-PPO-PEO and CiEOj surfactants has been extensively studied during the last decades. The interest in this topic has its origin in the scientific field but also in industrial applications of, e.g., the PEO-PPO-PEO copolymers,1-3 where such amphiphiles are used as detergents and emulsifiers and for controlled delivery purposes in pharmaceutical products.4,5 Due to their synergistic behavior, PEOPPO-PEO copolymers and surfactantssboth ionic and nonionics are often mixed together in solutions. There are numerous studies in the literature focusing on the interactions between various kinds of triblock copolymers mixed with ionic surfactants in water.6-19 Fewer studies have been published on mixtures with nonionic surfactants, see e.g., refs 20, 18, and 21. In previous investigations, our group has explored dilute solutions of the EO20PO68EO20 copolymer P123 and the nonionic * To whom correspondence should be addressed. E-mail:
[email protected]. Fax: +46 46222 4413. Tel.: +46 46222 1439. † Lund University. ‡ Present address: Dyrup A/S, Gladsaxevej 300, DK-2860 Soeborg, Denmark. § University of Graz. | University of Ljubljana.
surfactant C12EO6 and determined their interaction as well as the formation of mixed micelles by means of calorimetry and light scattering techniques.22,23 These studies led to the observation of a new phenomenon, namely, that the mixed micelles underwent a sphere-to-rod transition at a well-defined temperature that varied according to the C12EO6/P123 molar ratio.22 The shape transition temperature presented a minimum at a molar ratio between 2 and 3 and followed the same trend as the phase separation temperature at higher temperature. This shape transition was attributed to the PEO conformation change in the corona induced by the surfactants. Additionally, when the molar ratio was increased above 3, the spherical micelles decreased in size and the sphere-to-rod transition temperature increased again. This was explained by the larger amount of energy required by the system, due to the higher curvature of the smaller micelles, in order to change the micellar shape. This was also confirmed in a rheological study of the same system.24 Recently, the investigation has also been extended to include higher C12EO6/P123 molar ratios in order to understand the effect of the composition on the micellar structure. Here, the experimental techniques were static light scattering (SLS) and dynamic light scattering (DLS),23 and we found a compositioninduced structural change at constant temperature, which was manifested in the intermicellar interaction. The mixed micelles
10.1021/jp808442d CCC: $40.75 2009 American Chemical Society Published on Web 03/31/2009
Structural Characterization of Mixed Micelles changed from being spherical at low C12EO6/P123 molar ratios to polymer-like (“wormlike”) at high molar ratios. The present work aims at further investigating the shape of the P123-C12EO6 mixed micelles in the low molar ratio regime, where the temperature-induced sphere-to-rod transition occurs. This is done with the objective to understand the process of transition from a molecular point of view and to confirm the rodlike character of the mixed micelles after the transition. Various C12EO6/P123 molar ratios have been investigated at different total concentrations, with an emphasis on a molar ratio of 2.2. Three techniques have been used in order to characterize the P123-C12EO6 mixed micellar system; a DMA densiometer (density measurement apparatus) was utilized to achieve information concerning the viscosity in the solutions; the rotational motion of the rodlike mixed micelles was characterized by depolarized (VH) dynamic light scattering enabling a determination of the length; and depolarized scattering intensity (IVH) measurements were also performed as functions of temperature for various C12EO6/P123 molar ratios. Small-angle X-ray scattering (SAXS) was the main technique employed in order to characterize the shape and the physical properties of the mixed micelles, and indirect Fourier transformation (IFT) and generalized indirect Fourier transformation (GIFT) techniques were used to analyze the SAXS data. Experimental Section Materials. The PEO-PPO-PEO triblock copolymer (denoted P123) has an average composition of EO20PO68EO20 and a nominal molar mass of 5750 g mol-1.25 The sample was generously donated by BASF Corporation, Performance Chemicals, Mount Olive, NJ, and was used without further treatment. It is well-known that the PEO-PPO-PEO copolymers are polydisperse both in molar mass and composition,13,26-28 and recently a study of the effect of diblock impurities on the structural properties of ordered PEO-PPO-PEO micellar phases has been reported.29 The critical micelle concentration (cmc) of P123 may vary between different sample batches of different polydispersity and the experimental technique used. We have previously estimated the cmc at 40 °C to be 0.2 × 10-3 wt % (or 3.47 × 10-7 mol L-1) by using isothermal titration calorimetry (ITC)11,13 and 1.8 × 10-6 mol L-1 by using steadystate fluorescence spectroscopy.30 Wanka et al. has reported a value of 5.2 × 10-7 mol L-1.31 For a 1.0 wt % P123 solution, the critical micelle temperature is 17 ( 1 °C and the liquid-liquid phase separation temperature is 79 °C according to DSC experiments.22 Hexa(ethylene glycol) monododecyl ether (C12EO6) was obtained from Nikko Chemicals Co., Tokyo, Japan, and used as received. Its molar mass was 450.7 g mol-1. According to the gas chromatogram provided by the manufacturer, the purity of the sample was superior to 99%. The cmc of C12EO6 at 40 °C is 5.3 × 10-5 mol L-1 as determined using ITC,32 and the critical point for the liquid-liquid phase separation in a 2.25 wt % C12EO6 solution is about 47 °C.33 Water purified by a Milli-Q system (Millipore Corporation, Bedford, MA) was used in all solutions. The stock solutions of the P123 and C12EO6 were prepared by weighing and left to equilibrate overnight in a refrigerator at 4 °C. The mixed micellar solutions were prepared at various C12EO6/P123 molar ratios defined as the ratio between the number of moles of C12EO6 and P123. All samples were equilibrated at room temperature prior to the measurements. Ultralow Shear Viscosimetry. The viscosities of the micellar solutions were determined by utilizing a DMA 5000 (Anton Paar KG, Graz, Austria). This instrument consists of an
J. Phys. Chem. B, Vol. 113, No. 16, 2009 5479 oscillating tube and is principally a density meter. The effects from the viscosity lead to certain errors in the density measurements, and the measured densities thus have to be corrected for these effects. The sample viscosity causes a damping of the oscillating tube, and the corresponding signals are used for these corrections. Furthermore, the signals can be converted so as to become viscosity-dependent (a detailed description is given in ref 34) and thereby render the possibility of monitoring the changes in the sample viscosity at extremely low shear rates. All the P123-C12EO6 solutions that were studied with this technique were prepared to 5.0 wt % with respect to P123; i.e., the concentration of P123 [cP123 ) mP123/(mP123 + mC12EO6 + mwater) × 100 wt %] was kept constant. Approximately 1.0 mL of the solution was injected into the DMA instrument, whereupon the temperature was set. Measurements were performed on a series of temperatures with increments of 1 °C, and, in order to let the system reach thermal equilibrium, the measurements started 30 min after the set temperature for the scanning procedure was reached. The starting temperature was chosen to 25 °C for all measurements, whereas the final temperature varied from 50 to 75 °C depending on the molar ratio so as to cover the transition temperature in each case. Depolarized Light Scattering. Both static and dynamic depolarized light scattering (VH-SLS and VH-DLS) measurements were carried out using a laboratory-built precision laser goniometer (Institute of Chemistry, University of Graz, Austria), equipped with a 5 W single frequency solid-state diode-pumped laser with a wavelength λ ) 532 nm, Coherent, model Verdi, and two, high quality, Glan-Thompson polarizers (Halle, Berlin, Germany) with extinction ratios >10-6. The cylindrical light scattering cells (cuvettes with inner diameters of 8 mm) were immersed in a refractive-index-matching liquid (cisdecahydronaphthalene or decaline) kept in a cylindrical quartz container (VAT). The first polarizer was located in front of the VAT, and the second (the analyzer) in front of the detector. In order to measure the depolarized light scattering, the second polarizer was rotated 90° relative to the first one. The laser operated at an effective power of 1 W. The VH-SLS measurements were performed at the scattering angle θ ) 90° relative to the direction of the primary beam, whereas the VH-DLS experiments used angles from 35° to 135° with a step width of 5°. In the case of VH-SLS, the depolarized scattering intensity signal was achieved by subtracting the background intensity from the measured intensity. The background intensity was gained by measuring the signal without applying the laser light to the sample. In the case of VH-DLS, the analyzer was adjusted manually by a small rotation at each angle in order to find the optimum intensity of the anisotropic contribution (no VV contribution). Single-mode fiber optics were used (OZ, from GMP, Zurich, Switzerland) coupled to an ALV/SO-SIPD/DUAL photomultiplier with a pseudo-cross-correlation setup, ALV 5000 multiple tau digital correlator (ALV, Langen, Germany). The correlator constructed the intensity autocorrelation function (2) (q,t) of the depolarized scattered light. The regularized gVH inverse Laplace transformation fitting procedure REPES (regularized positive exponential sum),35,36 which directly minimizes the sum of the squared differences between the experimental (2) (q,t) and is incorporated in the GENDIST and calculated gVH package,37,38 was employed in the data analysis. The model used is expressed with respect to the normalized depolarized electric (1) (q,t), which in turn is related to field correlation function gVH (2) gVH(q,t) by Siegert’s relation (2) (1) gVH (q, t) - 1 ) β|gVH (q, t)| 2
(1)
5480 J. Phys. Chem. B, Vol. 113, No. 16, 2009
Lo¨f et al.
Here, t is the lag time and β is the coherence factor (e1) that takes into account the deviations from the ideal experimental situation (perfect coherence). q is magnitude of the scattering vector, q ) (4πn0/λ) sin(θ/2), in which n0 is the refractive index of the medium, i.e., solvent (in this study water) for dilute solutions. In a depolarized DLS experiment, the electric field correlation function at low q is a single-exponential function: (1) gVH (q, t) ) A exp(-ΓVHt) + baseline
(2)
Here, ΓVH is the depolarized relaxation frequency that consists of two parts: one that represents the translational diffusion D and another that represents the rotational diffusion Θ:
ΓVH ) 6Θ + q2D
(3)
ΓVH was extracted from the single-modal relaxation time distribution, which was obtained from the regularized inverse Laplace transformation of the depolarized intensity correlation (2) (q,t) measured at various angles. By plotting ΓVH functions gVH as a function of q2, the rotational diffusion coefficient was then obtained directly in the low-angle limit, qf0. Single-exponential Cumulant fits were also applied, and similar values of ΓVH were obtained. However, the Laplace inversion data showed better regression values for the fits of ΓVH vs q2, and these data are thus presented in this work. For rigid rodlike particles, the length L of the rods may be estimated from Θ by using an extended version39 of Broersma’s relation,40 but reformulated by Tirado et al:41
Θ)
3kT (ln p + δ) πη0L3
(4)
where p ) L/d, in which d is the rod diameter and η0 is the viscosity of the solvent (here water) and
δ ) -0.446 -
63 62 0.2 16 + 2 3 ln 2p (ln 2p) (ln 2p) (ln 2p)4 (5)
Small-Angle X-ray scattering. An evacuated high-performance SAXS instrument called “SAXSess” (Anton Paar KG, Graz, Austria) was used in this investigation. This instrument is a modification42 of the so-called “Kratky compact camera”.43 The SAXSess is attached to a conventional X-ray generator (Philips, The Netherlands) equipped with a sealed X-ray tube (Cu anode target producing Cu KR X-rays with a wavelength of 0.154 nm) operating at 40 kV and 50 mA. The samples were measured in a standard quartz capillary for the SAXSess camera (with an outer diameter of 1 mm and a wall thickness of 10 µm). The 2D scattering pattern was detected by a PI-SCX fused fiber optic taper charge-coupled device (CCD) camera from Princeton Instruments, a division of Roper Scientific, Inc. (Trenton, NJ). The used CCD detector featured a 2084 × 2084 pixel array with 24 µm × 24 µm pixel size (chip size: 50 mm × 50 mm) and was operated at -30 °C with a water-assisted cooling at 10 °C in order to reduce the thermally generated charge. Cosmic ray correction and background subtraction were performed on the 2D image before further data processing. The 2D image was integrated into the 1D scattering function I(q),
corrected for the solvent-filled capillary scattering and put to an absolute scale using water as a secondary standard.44 The obtained SAXS spectra were further analyzed with the indirect Fourier transformation method IFT45-47 as well as its extension, the generalized indirect Fourier transformation method GIFT.48-50 IFT is a model-free method used for dilute particle systems with negligible particle interactions (see description in refs 45-47). For systems with nonnegligible interparticle interactions (higher concentration of the scattering particles), GIFT was employed. Assuming a dilute system of globular particles with n scattering particles per volume element, the scattered intensity can be described by the product I(q) ) nP(q)S(q), where P(q) is the form factor representing the intraparticle scattering contributions, thus providing us with information of the particle size and shape, and S(q) is the structure factor representing the interparticle contributions. In dilute solutions, where interparticle interactions can be neglected, S(q) ) 1; i.e., the intensity is proportional to P(q). The scattering intensity from one scattering particle I1(q) can be written as the Fourier transformation of the so-called pairdistance distribution function p(r) describing the geometry of the scattering particles in the real space:46
I1(q) ) 4π
dr ∫0∞ p(r) sin(qr) qr
(6)
Here, r is the distance between two scattering centers within the particle. The p(r) function represents a histogram of the distance inside the scattering particle,48 which means that the function adopts a value of zero at a distance r greater than the maximum dimension of the particle. Particle size and shape are therefore two important parameters that can be obtained from the p(r) function. At higher concentration, where interparticle interactions have to be considered, the structure factor S(q) plays an important role and contributes to the total scattered intensity. Therefore, the interparticle correlation must be taken into account during the evaluation procedure. This can be handled with the GIFT technique.49-51 In this technique the interparticle interactions are described, in the real (r) space, by the total correlation function h(r) ) g(r) - 1, where g(r) is the radial distribution function and r is now defined as the distance between the centers of two particles.52 As for the form factor, the connection between the q- and r-space is a Fourier transform. S(q) and h(r) thus form the Fourier transform pair
S(q) ) 1 + 4πn
dr ∫0∞ h(r)r2 sin(qr) qr
(7)
S(q) can be described by a variety of models appropriate for the system under investigation. We determine P(q), which remains model-free, simultaneously with the model-dependent S(q) compensating for the interaction effects. The simplest structure factor model that takes into account the excluded volume effect for uncharged systems (“hard sphere model”) as well as the particle polydispersity is the average structure factor Save(q). This model is described by three parameters: volume fraction φ, interaction radius RHS (radius of hard spheres), and polydispersity µ. The effective volume fraction not only takes into account the surfactants and the copolymers but also includes the bound hydrated water. For a fixed concentration, this effective volume fraction will depend on the temperature because of the different hydration states. The interaction radius
Structural Characterization of Mixed Micelles
Figure 1. Relative viscosity (η/η0, where η0 is the water viscosity) as function of temperature for a P123-C12EO6 mixed micellar system at a C12EO6/P123 molar ratio of 2.2 (— —), 3.2 (– – –), 6.0 (– - –), and 11 (-...-), as well as for a pure C12EO6 solution (s). The P123 concentration in the mixed solutions was kept constant at 5.0 wt %.
in the hard sphere model (excluded volume interaction) is close to half of the minimum center-to-center distance of the particles. The polydispersity is defined as the ratio between the width of the Gaussian size distribution σ and the particle size RHS at its maximum value, i.e., µ ) σ/RHS.49 Results and Discussion Ultralow Shear Viscosimetry. Figure 1 displays how the viscosity, measured using the DMA instrument, of five aqueous solutions of P123-C12EO6 mixed micelles varied with the temperature. The values have been transferred to relative values (η/η0) by using literature viscosity values of water. Four C12EO6/ P123 molar ratios were investigated, 2.2, 3.2, 6.0, and 11, as well as a pure C12EO6 solution at 5.0 wt %. The P123 concentration in the mixed solutions was kept constant at 5.0 wt %. For each of the investigated molar ratios, the solution viscosity η initially displayed a slight decrease with temperature (
