The Nature of Fatty Acid Interaction with a Polyelectrolyte–Surfactant

Article pubs.acs.org/Langmuir

The Nature of Fatty Acid Interaction with a Polyelectrolyte− Surfactant Pair Revealed by NMR Spectroscopy Jose Martinez-Santiago,† Christian Totland,‡ Kavssery P. Ananthapadmanabhan,§ Liang Tsaur,§ and Ponisseril Somasundaran*,† †

NSF I/UCRC Center for Particulates and Surfactant Systems (CPaSS), Columbia University, New York, New York 10027, United States ‡ Department of Chemistry, University of Bergen, Allegaten 41, N-5007 Bergen, Norway § Unilever Research and Development, Trumbull, Connecticut 06611, United States S Supporting Information *

ABSTRACT: The interaction mechanisms of an oppositely charged polyelectrolyte−surfactant pair and dodecanoic (lauric) acid (LA) were experimentally investigated using a combination of nuclear magnetic resonance (NMR) techniques. It is observed that LA significantly affects the interaction between the anionic surfactant sodium dodecylethersulfate (SDES) and the cationic polymer guar modified with grafted hydroxypropyl trimethylammonium chloride (Jaguar C13 BF). Typically, oppositely charged polymers and surfactants interact electrostatically at a certain surfactant concentration known as the critical aggregation concentration (CAC). Once the polymer is neutralized by the surfactant, an insoluble complex (precipitate) is observed (phase separation), and, at concentrations beyond the surfactant critical micellar concentration (CMC′), the system returns to a one phase entity. In a system in which a mixture of SDES−LA is added to the polymer, NMR data show that below the neutralization onset, some of the polymer interacts with SDES, while some of the polymer is adsorbed at the surface of LA solid aggregates present in the system. Furthermore, SDES is found to aggregate in a lamellar-like structure at the polymer side chain prior to the SDES CMC′. Above the SDES (CMC′), LA is solubilized and incorporated at the palisade region of SDES micelles. Analysis of 1H resonances provided estimated concentrations of all species in the system phases at all stages of interaction.

1. INTRODUCTION

When looking at the phase behavior of polymer−surfactant systems, typically three steps are observed when adding a surfactant to an oppositely charged polyelectrolyte solution:2 (i) at a certain surfactant concentration known as the critical aggregation concentration (CAC) or S1, the surfactant starts interacting with the polymer via charge neutralization, and is progressively transformed into a more hydrophobic entity or complex; (ii) when the polymer is entirely neutralized by the surfactant, such complex becomes insoluble and a phase separation is observed. The onset of this phase separation is termed S1′; (iii) finally, near the surfactant critical micellar concentration (CMC′) or S2, a one-phase system is progressively observed. The CMC′ was traditionally associated with the onset of resolubilization of the polymer−surfactant complex by surfactant micelles, but a recent study by dos Santos et. al has shown that the resolubilization onset does not always occur at the CMC′ or S2.28 In this paper, the terms CMC′, S2, or resolubilization onset are used to designate the

The interaction mechanisms among polymers and surfactants in solution have been studied and reviewed for more than five decades.1−7 However, isolated polymer−surfactant systems rarely exist; most biological and synthetic systems containing polymers and surfactants are complex and possess a multicomponent nature. Literature in which a third component is included is emerging: recent studies focus on polymer− surfactant interactions in the presence of solid surfaces such as mica, alumina,8−11 and silica particles12−16 or, at the air/ water interface.17 Other interesting recent studies deal with the effect of an additional equivalent molecular component such as salt,18−21 an additional polymer, protein, or a mixture of surfactants.22−27 In this paper, the effects of dodecanoic (lauric) acid (LA) on the interaction of an oppositely charged polyelectrolyte− surfactant system composed of guar grafted with hydroxypropyl trimethylammonium chloride (Jaguar C13 BF polymer) and sodium dodecylethersulfate (SDES) surfactant was systematically studied using various 1H nuclear magnetic resonance (NMR) experiments. © 2014 American Chemical Society

Received: May 28, 2014 Revised: August 5, 2014 Published: August 11, 2014 10197

dx.doi.org/10.1021/la5020708 | Langmuir 2014, 30, 10197−10205

Langmuir

Article

concentration at which the system returns to a one phase entity, not necessarily the exact concentration at which the complex is resolubilized. The different stages of interaction among oppositely charged polymer−surfactant pairs were determined in the past using techniques such as surface tensiometry, fluorescent dyes, viscosity, and phase behavior,29 and to some extent, NMR. NMR spectroscopy can be applied to monitor chemical shift changes as a function of a solution composition and can provide molecular dynamics information, not accessible from other techniques, using tools such as NMR spin relaxation and molecular self-diffusion studies.6 For polymer−surfactant systems, NMR has previously shown significant potential for deducing the interaction of polymer− surfactant pairs;30−33 however, the vast potential of this analytical technique remains to some extent underutilized in such systems.34 In this paper, the phase behavior of a complex system composed of a polymer−surfactant pair in the presence of a third component (LA) is systematically studied by analyzing the 1H chemical shifts, 1H line widths, T1 relaxation measurements, nuclear Overhauser effect spectroscopy (NOESY) and diffusion ordered spectroscopy (DOSY). The unique combination of these techniques provided an excellent overview of not only the mechanisms of interaction of the polymer−surfactant system but also aided in the understanding of these systems in a multicomponent environment. The paper is organized as follows: in sections 3.1 and 3.2, the chemical shifts and T1 relaxation measurements are respectively presented and discussed. In section 3.3, NOESY results are shown. In section 3.4, analysis of the 1H resonances was performed in order to obtain the concentration of all components in the system at all stages of interaction. The polymer, SDES, and LA interaction mechanisms are schematically represented in Figure 9 at the end of section 3.4. Finally, the paper ends with the concluding remarks given in section 4.

Figure 1. Molecular structures of the Jaguar polymer, LA, and SDES. The protons are labeled according to the spectral assignments of Figure 2. (always adding the surfactant to the polymer while mixing). When LA was used, it was first mixed with the surfactant; the mixture was heated to 65 °C, stirred for 15 min, and cooled down to room temperature while stirring. Different concentrations of the SDES/LA mixture (at a constant SDES:LA molar ratio of 4:1) was then added to the polymer. The pH of the samples varied between 6.3 and 6.5 depending on the composition. 2.2.2. Dynamic Light Scattering (DLS) (Supporting Information). The average hydrodynamic diameter of particles was measured by DLS anemometry using a Zetasizer Nano ZS instrument (Malvern). The particles were placed in the disposable capillary cell for light scattering measurements. The Zetasizer detected the intensity of backscattered photons at a 173° angle from an incident 4 mW He−Ne (633 nm) laser over a time interval of 10 seconds, using a sample time (τ) of 0.5 μs. The hydrodynamic diameter was then extracted from the light scattering data using the method of cumulants.38 2.2.3. NMR Measurements. All NMR experiments were recorded at 300 K and carried out on a Bruker 500 Ultrashield spectrometer operating at 500 MHz for protons, except for the NOESY experiments which were carried out on a Bruker 600 Ultrashield spectrometer fitted with a cryo-probe for better resolution. D2O was used as solvent in all NMR experiments. The 1H chemical shifts were calibrated internally by setting the water proton resonance to 4.75 ppm. Line widths (supporting data) were measured by line shape deconvolution of the respective resonances using the Topspin (version 1.3) software. DOSY. (Supporting Information) A DOSY pulse scheme with trapezoidal-shaped gradient pulses was used. The diffusion times used range from 0.1 to 0.6 sec depending on the size and T1 relaxation times of the molecular components in question, and the gradient pulse length was typically 1-2.5 ms. The gradient pulse strength was varied in 16 linear steps in from 2 to 98% of the maximum gradient output. A delay time between scans of 5 seconds was used in all DOSY experiments. The signal attenuation was fitted using the equation

2. MATERIALS AND METHODS 2.1. Materials. The polymer used was hydrophilic polyelectrolyte guar modified with grafted hydroxypropyl trimethylammonium chloride (Jaguar C13S BF, Rhodia, Inc.; see Figure 1 for molecular structure). Its molecular weight is 1.8 × 106 g/mol and its nitrogen content ranges between 1.3 and 1.6%. The polyelectrolyte was purified using centrifugation at 7000 rpm for 30 min in order to remove small, water-soluble molecules such as salts and water-insoluble materials and proteins. The resulting concentration of purified polymer was then measured by using % solids in the system. Other physical properties of Jaguar C13S BF polymer in solution are described in detail elsewhere.35 The anionic surfactant used was SDES with one to three ethylene oxide groups, manufactured by Stepan Co. under the commercial name STEOL CS-170 (70% solution in water). SDES was used as received, without additional purification. The molecular mass of SDES is 332.4 g/mol (for 1EO). The CMC of unpurified SDES is ∼0.79 mM which was determined by surface tensiometry and is similar to other results found in the literature.36 The reported CMC value of purified SDES is significantly higher.37 Lauric (dodecanoic) acid (LA) >99.5% pure from Acros Organics was used as received, without additional purification. All solutions were prepared with deuterium oxide (D2O) of 99.8% purity from Cambridge Isotope Laboratories, Inc. 2.2. Methods. 2.2.1. Sample Preparation. For polymer− surfactant solutions, 0.1% weight polymer was first incorporated in the D2O solution with the help of an overhead mixer. The pH was increased above 9.0 using a 10% weight sodium hydroxide (D2O solution). After mixing for 10 min, the pH was decreased to 6.5 with 35% citric acid (D2O solution). Samples containing different SDES concentrations and a fix polymer concentration were then formulated

I = I0e−Dγ

2 2 2

g δ (Δ− δ /3)

(1)

where D is the diffusion coefficient, I is the observed signal intensity, I0 is the reference intensity (unattenuated signal intensity), Δ is the diffusion time, δ is the gradient pulse length, g is the gradient pulse 10198

dx.doi.org/10.1021/la5020708 | Langmuir 2014, 30, 10197−10205

Langmuir

Article

strength, and γ is the gyromagnetic ratio of proton. The signal attenuation plots the various system components are presented in the Supporting Information. NOESY. A zero-quantum suppressed pulse scheme was used. Water suppression using excitation sculpting with gradients was used due to difficulties in obtaining a sufficiently homogeneous field when LA solid aggregates were present, leading to large noise bands from the water proton resonance. The mixing time used for the displayed NOESY spectra is 0.5 sec. This value was set on the basis of several NOESY spectra recorded with variable mixing time between 0.1 and 0.8 sec, and was found to maximize the NOE, while at the same time obeying the linear approximation to avoid spin diffusion. This value is slightly below the T1 relaxation time of the fastest relaxing molecular component measured in the systems. 128 scans were acquired to obtain a resolution with acceptable signal-to-noise ratios. T1 Relaxation. A standard inversion recovery pulse scheme was used. Fifteen values of the delay time, τ, was used in each experiment, varying from 0.001 to 15 s. The delay time between each scan was set to 5 s.

SDES:LA molar ratio 4:1), the L2 CH2 resonance does not overlap with any SDES or polymer resonances, and this resonance is therefore the main focus regarding LA. It should be noted that even though the S7-15 and L4-11 resonances overlap, L4-11 only constitutes 20% of the total signal intensity; hence, the chemical shift of this resonance was taken as the S715 chemical shift in the below discussion. 3.1. Chemical Shifts Variations. Variations in 1H chemical shifts reflect changes in the electronic density near the hydrogen nucleus and are therefore useful for observing changes in the chemical environment around a molecule. In this section, chemical shift variations were analyzed for the S715 SDES resonance and the L2 LA resonance in different environments as a function of SDES and SDES/LA mixture concentration. In the absence of polymer, an increase in 1H chemical shifts was observed at concentrations above 0.79 mM of SDES, which correlates with its reported CMC (Figure 3A, ■). For surfactants, the observed chemical shift is normally a weighted average between bound/aggregated and nonbound molecules. The increase in chemical shift of the hydrocarbon tail protons upon micellation indicates a less crowded molecular environment in the micelle core. This is likely due to the exclusion of water molecules from the aggregate core, which will pack tightly around hydrocarbon moieties of nonaggregated surfactant monomers in bulk water.39 In the presence of LA (□), SDES shows a similar 1H chemical shift variation pattern but with slightly increased values, especially after the CMC. This reflects a change in molecular packing upon SDES-LA mixed micelle formation, and may be caused by an increase of C−C trans conformers in the SDES hydrocarbon tail when LA is present in the micelle. The formation of mixed micelles is confirmed by a significant increase in the LA L2 1H chemical shift above the SDES CMC (Figure 3A, ■). Figure 3 (B) shows 1H chemical shift variations of SDES (S715) in the presence of polymer, with (◇) and without (◆) LA. Three different stages of interaction are observed when SDES binds to the oppositely charged Jaguar polymer, represented by the three characteristic SDES concentrations S1 (CAC), S1′, and S2 or CMC′ (indicated by a prime due to the delay in CMC when polymer is present). At S1, or the CAC, the surfactant starts binding to the feeble surface active polymer via electrostatic interaction, which is transformed into a more hydrophobic entity or complex. At S1′, the polymer is entirely neutralized by the surfactant, the complex becomes insoluble, and a precipitate is visually observed. Finally, at S2, or CMC′, surfactant micelles form, and a one-phase system is progressively observed. In Figure 3 (right), SDES 1H chemical shift variations show the three stages of interaction; from S1 to S1′, SDES binds to the polymer, causing a slight increase in the 1 H chemical shift. The gradual increase in chemical shift at this point indicates some degree of aggregation of SDES at the polymer chain. At S1′, precipitation takes place and the observed 1H signals thus come from the SDES that remains in the supernatant. The observed decrease in chemical shift above S1′ is therefore due to an increasing precipitation of polymer which causes nanoaggregated SDES molecules to contribute more to the observed chemical shift. Beyond S2, the chemical shifts of the resolubilized complex are measured, which change to higher values due to the formation of micelles. Figure 3 also shows chemical shift variations of systems where the SDES/LA mixture is added to the polymer (◇). When LA is present, the

3. RESULTS AND DISCUSSION Figure 2 shows the 1H NMR spectra of an SDES/LA mixture (top) and a Polymer/SDES/LA mixture (bottom) with the

Figure 2. 1H spectra of 50 mM SDES/12.5 mM LA (top) and 0.79 mM SDES/0.2 mM LA/0.1 wt % polymer (bottom) in D2O solution at 300 K. The HDO peak increases in width and intensity when polymer is present due to increased viscosity and some SDES H2O content, respectively. The peaks are partly assigned based on 1H−13C HSQC (Supporting Information) and NOESY spectra and labeled according to the labels given in Figure 1.

resonances partly assigned. From Figure 2 (bottom), it is apparent that some polymer 1H resonances overlap with the S1-4 1H SDES resonances. However, the S7-15, S16 and S5-6 SDES resonances, as well as the nitrogen bound polymer CH3 (Pcat) and side chain resonances (Psc), do not overlap. Results in this study therefore focus mostly on these resonances, with the exception of the Pcat resonance which displayed some unfavorable characteristics due to fast spin-rotation about the H3C−N bond. Furthermore, when LA is present (always at 10199

dx.doi.org/10.1021/la5020708 | Langmuir 2014, 30, 10197−10205

Langmuir

Article

Figure 3. (A) 1H chemical shifts of SDES (S7-15) in D2O solution with (□) and without (■) LA, and 1H chemical shifts of the LA L2 resonance (light grey ■) in a SDES/LA D2O solution. (B) 1H chemical shift variations of SDES (S7-15) in 0.1% w/w of polymer with (◇) and without (◆) LA. The chemical shifts are recorded in D2O at 300 K. All solutions containing LA have an SDES:LA 4:1 molar ratio. Uncertainty is within the size of the symbols used.

existence of LA aggregates in the bulk (Supporting Information). 3.2. T1 Relaxation Measurements. NMR spin−lattice relaxation (T1) is frequently used to monitor dynamical properties of molecules in liquid solution due to the high sensitivity of the dipolar relaxation mechanism to molecular mobility. T1 relaxation is affected by fast molecular motions and hence reflects local motions of a molecule, rather than, for example, the motion of an aggregated group of molecules such as a micelle or a polymer−surfactant complex. Therefore, in nonreactive multimolecular systems, a decrease in T1 relaxation for a molecule is normally associated with binding, adsorption, or proximity to another molecule which lowers its motional freedom. T1 relaxation was measured for SDES/LA mixtures at concentrations below and above the SDES CMC (keeping the SDES:LA molar ratio 4:1) with and without the presence of 0.1% weight polymer. Figure 5A (□) shows that the SDES hydrocarbon chain mobility decreases slightly after the CAC which is at ∼0.097 mM of SDES. A decrease in mobility, in this case, means that SDES monomers are binding to the polymer. When precipitation occurs, the bound SDES are no longer present in the supernatant and, thus, do not contribute to the recorded NMR SDES resonance. Therefore, at S1′ (precipitation onset), the T1 relaxation values of SDES increase due to a shift in the equilibrium between free and bound SDES molecules toward the more mobile-free SDES. The precipitation onset was also observed at the same concentration (0.4 mM of SDES) when measuring the chemical shift variations (section 3.1). At S2, SDES mobility and T1 relaxation values decrease as the SDES resonance is increasingly dominated by the resolubilized polymer-bound and micellized SDES. This phenomenon was also observed by the increase in chemical shift in section 3.1. As expected, due to the strong correlation between recorded T1 values and precipitation/resolubilization, the T1 data for the polymer side chain (Psc) show the same trends as the SDES T1 data for the polymer/SDES sample (Figure 5B, ◇). However, below the CMC′ (S2), it was necessary to assume two T1 relaxation components for Psc in order to get an acceptable fit to the inversion recovery decay curves for the polymer/SDES/ LA system (Figure 5B, ●). This means that, when LA is in the system, the polymer is present in two environments/states, which give rise to significantly different mobility for the polymer side chain, and the exchange between these environments/states is sufficiently slow for both to be observed. The

same three regions are observed: below S2, the SDES chemical shifts are slightly lower compared to the LA-free system. Above S2, SDES resolubilizes with higher chemical shift values compared to samples without LA, indicating the existence of LA/SDES mixed micelles with a slightly different molecular packing than the SDES micelle. This is discussed further in section 3.3. In Figure 4, the chemical shift variations of the LA L2 resonance are shown as a function of SDES concentration with

Figure 4. 1H chemical shift variations of LA L2 protons in D2O in the presence of SDES (▲, also given in Figure 3 (left) for comparison) and in the presence of SDES and 0.1% w/w polymer solution (△).

and without the presence of polymer (note that the SDES:LA molar ratio is 4 in all samples containing LA). In the polymerfree system (▲), the LA chemical shift values increased above SDES CMC, indicating LA solubilization into the SDES micelles. When the polymer is present (△), an increase in chemical shift values is observed above S2 (CMC′ or resolubilization onset), also due to mixed micelle formation. However, there are no chemical shift variations for the L2 resonance prior to the CMC′. This indicates that there is no significant interaction between LA and the polymer prior to the SDES CMC′. Below CMC′, LA concentration is sufficient for it to be mostly present at the bulk, with a marginal amount at the air/ water interface. Other results from this study (see, e.g., section 3.4) strongly suggest that the formation of solid LA aggregates or nanocrystals occurs. LA in solid aggregate form would not contribute to the observed LA resonance, which will represent the monomeric and micellized LA molecules present in the bulk. DLS run on samples below CMC′ confirmed the 10200

dx.doi.org/10.1021/la5020708 | Langmuir 2014, 30, 10197−10205

Langmuir

Article

interaction with the LA aggregates, which increases the local mobility of some of the side chains. Polymer adsorption to solid state LA was observed elsewhere40 for the same system, and it was associated with the increased negative surface charge associated with LA aggregates in solution. When LA is present (Figure 5A, ■), increased T1 relaxation values demonstrate that SDES mobility is higher before the precipitation onset compared to the LA-free system. This may be related to the high mobility of the polymer relaxation component observed in this sample. Assuming this component originates from polymer adsorbed at LA surfaces, there may be fewer accessible adsorption sites for the SDES molecules, hence more mobile SDES is present in the system. Because the recorded T1 values are weighted average values of free and adsorbed SDES, the reduction in adsorption sites could shift the observed T1 value in the direction of the more mobile-free SDES. Self-diffusion data obtained from DOSY experiments (Supporting Information) confirmed that there are fewer binding sites for SDES on the polymer when LA is present. Figure 5C shows the T1 relaxation of LA. In the absence of polymer, an inflection point is observed near the CMC of SDES due to LA incorporation/solubilization in the SDES micelles. In the presence of polymer (Figure 5, C ▲), the LA signal intensity was too low for a reasonable evaluation of the T1 relaxation at concentrations below and at the CAC, which made it more difficult to interpret the interaction pattern of LA close to precipitation. A slight decrease in mobility is observed before precipitation; however, the change in T1 relaxation before the resolubilization onset for the LA L2 protons of about -0.10 sec is by no means as pronounced as the +0.38 s difference observed for the SDES S7-15 protons, again indicating that LA bulk monomers are not interacting with the polymer to a significant extent at these concentrations. A large decrease in T1 relaxation values after S2 (CMC′) is observed due to the restricted mobility of LA in SDES micelles. In order to obtain information on slower molecular motions than those affecting the T1 relaxation, 1H resonance line widths and DOSY data were analyzed. DOSY results are included in the supporting data and agree with the proposed interaction mechanisms of SDES and SDES/LA mixtures with Jaguar polymer. So far, results showed that SDES interacts with the polymer above S1. At this stage, LA forms aggregates in the nanometer range which may induce adsorption of some of the polymer to the LA aggregates. From S1′ to S2, precipitation of SDESsaturated polymer takes place. Above S2, SDES micelles are capable of solubilizing LA aggregates, and, with an excess of SDES/LA mixed micelles, the system goes back to a one phase entity. To further investigate the molecular interactions of all species in the system, NOESY was performed. Moreover, the concentration of all molecules in the system at the different stages of interaction is discussed in section 3.4 to further elucidate the mechanisms of interaction of the various molecular components. 3.3. Nuclear Overhauser Effect Spectroscopy. The NOESY experiment is useful for determining which molecular segments are close to each other in space; protons (both of different molecules and within the same molecule) closer than about 5 Å will result in an off-diagonal cross-peak in the twodimensional NOESY spectrum. Here, the NOESY experiment was used to investigate how the surfactant interacts with the polymer, the effect of LA in the system, and the structure of

Figure 5. T1 as a function of SDES concentration for the S7-15 SDES resonance (A), the polymer side chain resonance (B), and the LA L2 resonance (C), in the SDES/polymer (□, ◇), SDES/LA/Polymer (■, ●, ▲), and SDES/LA (light grey ▲) systems.

low-mobility component relaxes 0.1−0.2 seconds faster, and the high-mobility component relaxes as much as 0.7 seconds slower than the single component observed in the Polymer/SDES sample. One plausible explanation for the high- and lowmobility component is that the presence of LA aggregates in the bulk may induce polymer adsorption. Adsorbed polymer to LA aggregates will show a different relaxation time compared to SDES-bound polymer; however, the additional relaxation component indicates polymer side chains with high motional freedom. In fact, the T1 relaxation time of about 1.5 seconds for the additional polymer component is more than twice as high as the value measured for polymer alone of 0.63 seconds. It is possible that a change in polymer conformation occurs upon 10201

dx.doi.org/10.1021/la5020708 | Langmuir 2014, 30, 10197−10205

Langmuir

Article

Figure 6. Left: NOESY spectrum of the SDES/LA in 0.1% w/w polymer in D2O well above S2. Right: NOESY spectrum of the SDES/LA in 0.1% w/w polymer in D2O close to S2. Both spectra are recorded under similar experimental conditions and displayed with similar spectral enhancement.

polymer side chain for the SDES molecules, which may be a driving force in the resolubilization process. It should be noted that no cross-peaks were expected between the Pcat resonance and any other molecular components due to fast bond-rotation about the H3C−N bonds. This resulted in T1 relaxation rates of more than 6 s, which makes it improbable that a sufficient NOE build-up can occur with the 0.5 s mixing time used in the displayed spectra. Second, well above S2, there are cross-peaks between the second methylene of LA (L3) and SDES headgroup protons (methylenes near the ether groups) closest to the sulfate group, demonstrating that LA is incorporated at the outer part of the SDES palisade layer (cross-peaks indicated by 1′ in Figure 6, left) where the L2 protons are sufficiently separated from the SDES protons to not give a cross-peak. None of these crosspeaks appear below S2, so, as concluded earlier, LA is not interacting to a significant extent with SDES until the CMC′. Furthermore, no cross-peaks are observed between any polymer resonances and the LA L2 resonance at any concentrations, further indicating that LA is solubilized by SDES micelles after the CMC′. Figure 7 illustrates the structure of SDES/LA mixed micelles based on the NOESY spectrum. 3.4. Concentration Analysis. Concentrations of SDES, LA, and polymer side chain at all stages of interaction could be determined by integrating the 1H resonances. For samples above S1′ (precipitation onset) and below S2 (one phase system), the resonances of the supernatant were analyzed. Figure 8B shows SDES and the polymer side chain concentrations at all stages of interaction. A value of 1.0 means that the measured concentration of the supernatant is the same as the total concentration of that substance in the sample, and a value below 1.0 means that some molecules have adsorbed/precipitated causing the supernatant concentration to be lower than the total concentration. As expected, below S1′, SDES (■) starts binding to the polymer, hence, its value is below 1.0. The increase of the values from 0.6 to 1.0 prior S1′ is

SDES aggregates at the polymer chain. Polymer−surfactant models have been proposed previously, all of which coincide in the formation of a necklace-like structure when surfactant binds to the polymer; however, no specific details were given regarding orientation of surfactant molecules at the polymer chain. The necklace model is, to date, the most accepted model for surfactant−polymer systems. Figure 6 shows NOESY spectra of the SDES/LA/polymer system at concentrations close to S2 (left) and well above S2 (right). Above S2, the system is composed of a complex formed of SDES/LA and polymer which is resolubilized by the excess of SDES/LA mixed micelles. Below S2, the presence of the anisotropic polyelectrolyte, as well as LA solid aggregates, resulted in some noise in the NOESY spectra; nevertheless, key cross-peaks are still clearly observable. There are two important outcomes from these spectra; first, some of the polymer side chain is close in space to the SDES carbon chain at concentrations below and close to S2 (crosspeak indicated by 2 in Figure 6, right). Furthermore, SDES aggregates are also observed close to S2; cross-peaks between the SDES headgroup and the CH3 protons (cross-peaks indicated by 1 in Figure 6, right) demonstrate that the SDES pack in an antiparallel conformation where adjacent molecules are aligned with the head groups in opposite directions. This indicates the formation of SDES premicellar aggregates in a lamellar-like structure at the polymer chain below CMC′. Such structures have been found previously in the insoluble complex of polyelectrolyte−surfactant systems using small-angle X-ray diffraction (SAXS),41 but evidence of such structures in the soluble bulk part of the system are scarce. At concentrations well above S2, there is no proximity of the SDES carbon chain to the polymer side chain, and no SDES lamellar aggregates are observed. This is an indication of a change in SDES configuration at the polymer chain: from lamellar aggregates to micelles. These results suggest that from S2 the micelle is a more preferable environment than the 10202

dx.doi.org/10.1021/la5020708 | Langmuir 2014, 30, 10197−10205

Langmuir

Article

complexes precipitate progressively and that less SDES is required for binding closer to S2 due to the proximity of the polymer to the saturation point. At S2, or CMC′, when the maximum precipitate amount is formed, Figure 8 shows that around half of the amount of SDES originally added to the sample is depleted from the supernatant and around 80% of the polymer precipitates. Above S2, resolubilization of the complex occurs progressively, and all components are back in to solution at ∼25 mM of SDES, which is designated as (S3) in Figure 8. The effect of LA is shown in Figure 8A. There are two important outcomes from this plot below the precipitation onset S1′: (i) the majority of LA does not contribute to the observed 1H resonance because it is aggregated in solid form and, (ii) contrary to the LA-free system (Figure 8B), a significant amount of polymer is not present in the bulk just below S1′. Despite the fact that a visual inspection does not reveal a precipitate at this stage, it is possible that, in the presence of LA aggregates, the precipitation onset S1′ may have occurred at an earlier stage in which polymer−surfactant nanoaggregates not detectable with a naked eye may have formed. Interestingly, turbidity measurements performed on the supernatant of these samples below, at, and above S1′ revealed an increased value in turbidity just below S1′. Chemical shift variations and T1 relaxation measurements in systems containing LA aggregates showed the presence of increased amounts of nonbound SDES below S1′ due to less available polymer to bind (also showed in Figure 8A); a smaller amount of polymer in the system will reach a neutralization point at an earlier stage. The adsorption of almost neutral polymer to an LA-rich air/water interface near the precipitation onset will slightly deplete the bulk of polymer without any visible precipitation and may be an alternative explanation. This is less likely due limited capacity of the air/water interface to host such a large amount of polymer. Between S1′ and S2, the SDES and polymer concentration decreases, confirming the composition of the complex formed of polymer and SDES. LA concentrations below S2 do not take into account the solid state LA, but only the few detectable LA monomers in the bulk. Figure 8A shows that the concentration of LA monomers decreases in the bulk from S1′ to S2. Since there are no cross-peaks (binding) between LA and SDES or the polymer before S2 (NOESY), it is hypothesized that, while polymer−surfactant complexes precipitate, some LA monomers may migrate to the existing LA nanoaggregates in the bulk, while some LA may cumulate at the air/water interface; consequently a decrease in the LA NMR signal is observed in the bulk from S1′ to S2. Surface tensiometry experiments showed how the surface tension kept decreasing from S1′ to S2 for the same system, indicating the accumulation of LA at the air/water interface before CMC′.40 Such phenomenon was not observed in the LA-free system. Another interesting observation is that, in systems with LA, slightly increased amounts of SDES and polymer migrate to the complex compared to the LA-free system. At the maximum precipitation stage, the complex contained ∼60% of SDES and almost 85% of polymer originally added to the system. Figure 9 shows a schematic representation of the mechanisms of interaction of the polymer, SDES, and LA system as a function of SDES/LA concentration. This model takes into account all our experimental findings, including surface tension measurements results published elsewhere,40 which relates to the surface component of this figure.

Figure 7. Schematic representation of a section of the SDES micelles with incorporated LA.

Figure 8. Total SDES concentration versus the ratio [supernatant]: [total] for the three components of the polymer/SDES/LA sample (A) and the two components of the polymer/SDES sample (B).

due to increasing the free surfactant in the dilute SDES concentration regime. The polymer supernatant concentration (●) remains the same until precipitation, which is expected, as the polymer concentration is kept constant for all samples. Above S1′, both polymer and SDES concentrations decrease in the supernatant at a similar rate. By looking at the depletion rate of free SDES and polymer side chain from the supernatant between S1′ and S2, we can conclude that SDES/polymer 10203

dx.doi.org/10.1021/la5020708 | Langmuir 2014, 30, 10197−10205

Langmuir

Article

Figure 9. Schematic representation of the Jaguar polymer side chain between S1 and S3. The proximity between the SDES tail and polymer sidechain and the lamellar SDES structure is indicated, as well as the position of LA in the palisade layer of the SDES micelle. For a better visual effect, the polymer chains are amplified in this figure (see the DLS data in Supporting Information for the actual size distribution of polymer and LA aggregates below S1′).

■

4. CONCLUSIONS The interaction mechanisms of polymer−surfactant1−7 and surfactant−fatty acid36,42 binary mixtures were extensively studied in the past. The fact that these materials coexist in numerous biological and industrial systems strongly motivated the study of a ternary system composed of polymer, surfactant, and fatty acid. In this paper we focused on the effects of dodecanoic acid on the well understood complexation mechanism of an oppositely charged polyelectrolyte−surfactant system. A systematic study using NMR techniques demonstrated that, below the precipitation onset, some of the polymer is progressively neutralized by the oppositely charged surfactant, which gradually forms lamellar aggregates at the polymer side chain, while some polymer is presumably adsorbed to the insoluble LA nanoaggregates present in the system. With increased amounts of SDES, when micelles form, LA transitions from solid-state aggregate to solubilized state at the palisade region of SDES micelles. Finally, it was observed that, in the presence of LA, the complexation efficiency of the polymer−surfactant pair increased. The new findings and the systematic NMR study presented in this paper are not only useful to study the incorporation of fatty acids in polymer−surfactant complexes but could also have important practical implications in a vast number of colloidal formulations where water-soluble polymers, surfactants, and fatty acids are present.

■

AUTHOR INFORMATION

Corresponding Author

*Mailing address: NSF I/UCRC Center for Particulates and Surfactant Systems, Columbia University, NY, New York, 10027, USA. Phone: 845 216 3543. Fax: 212 854 8362. E-mail: [email protected]. Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The financial support is partially from the National Science Foundation (NSF) under Grant No. 0749481 and by the industry members of the NSF I/UCRC Center for Particulates and Surfactant Systems (CPaSS) at Columbia University.

■

REFERENCES

(1) Ananthapadmanabhan, K. P.; Leung, P. S.; Goddard, E. D. Fluorescence and Solubilization Studies of PolymerSurfactant Systems. Colloids Surf. 1985, 13, 63−72. (2) Goddard, E. D. Principles of Polymer Science and Technology in Cosmetics and Personal Care; Marcel Dekker, Inc.: New York, 1999; Vol. 22. (3) Goddard, E. D.; Leung, P. S. Complexes of Cationic Polymers and Anionic Surfactants. In Polymer Science and Technology; Abstracts of Papers of the American Chemical Society; American Chemical Society: Washington, DC, 1985; 30, pp 407−415. (4) Goddard, E. D. PolymerSurfactant Interaction Part I. Uncharged Water-Soluble Polymers and Charged Surfactants. Colloids Surf. 1986, 19, 255−300. (5) Goddard, E. D. PolymerSurfactant Interaction Part II. Polymer and Surfactant of Opposite Charge. Colloids Surf. 1986, 19, 301−329.

ASSOCIATED CONTENT

S Supporting Information *

(1) 1H resonances line widths; (2) diffusion ordered spectroscopy (DOSY); (3) 1H−13C HSQC/1H−1H COSY; and (4) dynamic light scattering on a polymer/SDES/LA sample. This material is available free of charge via the Internet at http:// pubs.acs.org 10204

dx.doi.org/10.1021/la5020708 | Langmuir 2014, 30, 10197−10205

Langmuir

Article

(6) Kwak, J. C. T. Polymer−Surfactant Systems; Marcel Dekker, Inc.: New York, 1998; Vol. 77. (7) Lindman, B.; Thalberg, K. Interactions of Surfactants with Polymers and Proteins; CRC Press: Boca Raton FL, 1992. (8) Dedinaite, A.; Claesson, P. M.; Nygren, J.; Iliopoulos, I. Interactions Between Surfaces Coated with Cationic Hydrophobically Modified Polyelectrolyte in the Presence and the Absence of Oppositely Charged Surfactant. In Surface and Colloid Science; Razumas, V.; Lindman, B.; Nylander, T., Eds.; Springer-Verlag: Berlin, 2001; Vol. 116, pp 84−94. (9) Grządka, E. Competitive Adsorption in the System: Carboxymethylcellulose/Surfactant/Electrolyte/Al2O3. Cellulose 2011, 18, 291−308. (10) Omarova, K. I.; Musabekov, K. B.; Skachkova, A. L. Adsorbed Layers of Associates of Poly-Electrolytes with Nonionic Surfactants on Solid-Surfaces. Colloid J. 1989, 51, 144−149. (11) Qian, L. M.; Charlot, M.; Perez, E.; Luengo, G.; Potter, A.; Cazeneuve, C. Dynamic Friction by Polymer/Surfactant Mixtures Adsorbed on Surfaces. J. Phys. Chem. B 2004, 108, 18608−18614. (12) Carnali, J. O.; Shah, P. Correlation of Surfactant/Polymer Phase Behavior with Adsorption on Target Surfaces. J. Phys. Chem. B 2008, 112, 7171−7182. (13) Furst, E. M.; Pagac, E. S.; Tilton, R. D. Coadsorption of Polylysine and the Cationic Surfactant Cetyltrimethylammonium Bromide on Silica. Ind. Eng. Chem. Res. 1996, 35, 1566−1574. (14) Meagher, L.; Maurdev, G.; Gee, M. L. Interaction Forces Between a Bare Silica Surface and an α-Alumina Surface Bearing Adsorbed Polyelectrolyte and Surfactant. Langmuir 2002, 18, 2649− 2657. (15) Roques-Carmes, T.; Aouadj, S.; Filiatre, C.; Membrey, F.; Foissy, A. Interaction Between Poly(vinylimidazole) and Sodium Dodecyl Sulfate: Binding and Adsorption Properties at the Silica/ Water Interface. J. Colloid Interface Sci. 2004, 274, 421−432. (16) Yang, B.; Jaber, R.; Edler, K. J. Silica−Surfactant−Polyelectrolyte Film Formation: Evolution in the Subphase. Langmuir 2012, 28, 8337−8347. (17) Penfold, J.; Thomas, R. K.; Taylor, D. J. F. Polyelectrolyte/ Surfactant Mixtures at the Air−Solution Interface. Curr. Opin. Colloid Interface Sci. 2006, 11, 337−344. (18) Abraham, A.; Mezei, A.; Meszaros, R. The Effect of Salt on the Association Between Linear Cationic Polyelectrolytes and Sodium Dodecyl Sulfate. Soft Matter 2009, 5, 3718−3726. (19) Felippe, A. C.; Bellettini, I. C.; Eising, R.; Minatti, E.; Giacomelli, F. C. Supramolecular Complexes Formed by the Association of Poly(ethyleneimine) (PEI), Sodium Cholate (NaC) and Sodium Dodecyl Sulfate (SDS). J. Braz. Chem. Soc. 2011, 22, 1539−1548. (20) Pojjak, K.; Bertalanits, E.; Meszaros, R. Effect of Salt on the Equilibrium and Nonequilibrium Features of Polyelectrolyte/Surfactant Association. Langmuir 2011, 27, 9139−9147. (21) Wang, X. Y.; Li, Y. J.; Li, J. X.; Wang, J. B.; Wang, Y. L.; Guo, Z. X.; Yan, H. Salt Effect on the Complex Formation Between Polyelectrolyte and Oppositely Charged Surfactant in Aqueous Solution. J. Phys. Chem. B 2005, 109, 10807−10812. (22) Kizilay, E.; Kayitmazer, A. B.; Dubin, P. L. Complexation and Coacervation of Polyelectrolytes with Oppositely Charged Colloids. Adv. Colloid Interface Sci. 2011, 167, 24−37. (23) Liu, J.; Takisawa, N.; Shirahama, K. The Interaction of Mixed Surfactants with Polyelectrolytes. Colloid Polym. Sci. 1999, 277, 247− 251. (24) Penfold, J.; Tucker, I.; Thomas, R. K.; Taylor, D. J. F.; Zhang, J.; Bell, C. Influence of the Polyelectrolyte Poly(ethyleneimine) on the Adsorption of Surfactant Mixtures of Sodium Dodecyl Sulfate and Monododecyl Hexaethylene Glycol at the Air−Solution Interface. Langmuir 2006, 22, 8840−8849. (25) Staples, E.; Tucker, I.; Penfold, J.; Warren, N.; Thomas, R. K. Organization of Polymer−Surfactant Mixtures at the Air−Water Interface: Poly(dimethyldiallylammonium chloride), Sodium Dodecyl

Sulfate, and Hexaethylene Glycol Monododecyl Ether. Langmuir 2002, 18, 5139−5146. (26) Tajik, B.; Sohrabi, B.; Amani, R.; Hashemianzadeh, S. M. The Study of Polymer−Surfactant Interaction in Catanionic Surfactant Mixtures. Colloids Surf., A 2013, 436, 890−897. (27) Zhang, X. L.; Taylor, D. J. F.; Thomas, R. K.; Penfold, J. The Effects of the Addition of the Polyelectrolyte, Poly(ethyleneimine), on the Adsorption of Mixed Surfactants of Sodium Dodecylsulfate and Dodecyldimethylaminoacetate at the Air−Water Interface. J. Colloid Interface Sci. 2011, 356, 647−655. (28) dos Santos, S.; Gustavsson, C.; Gudmundsson, C.; Linse, P.; Piculell, L. When Do Water-Insoluble Polyion−Surfactant Ion Complex Salts “Redissolve” by Added Excess Surfactant? Langmuir 2010, 27, 592−603. (29) Goddard, E. D. Polymer/Surfactant Interaction. J. Soc. Cosmet. Chem. 1990, 41, 23−49. (30) Barhoum, S.; Yethiraj, A. An NMR Study of Macromolecular Aggregation in a Model Polymer−Surfactant Solution. J. Chem. Phys. 2010, 132, 9. (31) Denkova, P. S.; Van Lokeren, L.; Willem, R. Mixed Micelles of Triton X-100, Sodium Dodecyl Dioxyethylene Sulfate, and Synperonic L61 Investigated by NOESY and Diffusion Ordered NMR Spectroscopy. J. Phys. Chem. B 2009, 113, 6703−6709. (32) Pettersson, E.; Topgaard, D.; Stilbs, P.; Soderman, O. Surfactant/Nonionic Polymer Interaction. A NMR Diffusometry and NMR Electrophoretic Investigation. Langmuir 2004, 20, 1138−1143. (33) Sjogren, H.; Ericsson, C. A.; Evenas, J.; Ulvenlund, S. Interactions Between Charged Polypeptides and Nonionic Surfactants. Biophys. J. 2005, 89, 4219−4233. (34) Scheler, U. NMR on Polyelectrolytes. Curr. Opin. Colloid Interface Sci. 2009, 14, 212−215. (35) Gittings, M. R.; Cipelletti, L.; Trappe, V.; Weitz, D. A.; In, M.; Marques, C. Structure of Guar in Solutions of H2O and D2O: An Ultra-Small-Angle Light-Scattering Study. J. Phys. Chem. B 2000, 104, 4381−4386. (36) Tzocheva, S. S.; Kralchevsky, P. A.; Danov, K. D.; Georgieva, G. S.; Post, A. J.; Ananthapadmanabhan, K. P. Solubility Limits and Phase Diagrams for Fatty Acids in Anionic (SLES) and Zwitterionic (CAPB) Micellar Surfactant Solutions. J. Colloid Interface Sci. 2012, 369, 274− 86. (37) Xu, H.; Penfold, J.; Thomas, R. K.; Petkov, J. T.; Tucker, I.; Webster, J. P. R. The Formation of Surface Multilayers at the Air− Water Interface from Sodium Polyethylene Glycol Monoalkyl Ether Sulfate/AlCl3 Solutions: The Role of the Size of the Polyethylene Oxide Group. Langmuir 2013, 29, 11656−11666. (38) Frisken, B. J. Revisiting the Method of Cumulants for the Analysis of Dynamic Light-Scattering Data. Appl. Opt. 2001, 40, 4087−91. (39) Frank, H. S.; Evans, M. W. Free Volume and Entropy in Condensed Systems III. Entropy in Binary Liquid Mixtures; Partial Molal Entropy in Dilute Solutions; Structure and Thermodynamics in Aqueous Electrolytes. J. Chem. Phys. 1945, 13, 507−532. (40) Martinez-Santiago, J.; Ananthapadmanabhan, K..P.; Tsaur, L.; Totland, C.; Somasundaran, P. Effects of Fatty Acids on Polyelectrolyte-Surfactant Interactions. Implications for Polymer-Induced Flocculation/Dispersion in Emulsion Systems. Colloids Surf., A 2014, 461, 57−65. (41) Kogej, K. Association and Structure Formation in Oppositely Charged Polyelectrolyte-Surfactant Mixtures. Adv. Colloid Interface Sci. 2010, 158, 68−83. (42) Garti, N. Effect of Surfactants on Crystallization and Polymorphic Transformation of Fats and Fatty Acids; Marcel Dekker Inc.: New York, 1988; Vol. 31.

10205

dx.doi.org/10.1021/la5020708 | Langmuir 2014, 30, 10197−10205