Structure of Polystyrenesulfonate/Surfactant Mixtures at Air-Water Interfaces and Their Role as Building Blocks for Macroscopic Foam.

Air/water interfaces were modified by oppositely charged poly(sodium 4-styrenesulfonate) (NaPSS) and hexadecyltrimethylammonium bromide (CTAB) polyelectrolyte/surfactant mixtures and were studied on a molecular level with vibrational sum-frequency generation (SFG), tensiometry, surface dilatational rheology and ellipsometry. In order to deduce structure property relations, our results on the interfacial molecular structure and lateral interactions of PSS-/CTA+ complexes were compared to the stability and structure of macroscopic foam as well as to bulk properties. For that, the CTAB concentration was fixed to 0.1 mM, while the NaPSS concentration was varied. At NaPSS monomer concentrations <0.1 mM, PSS-/CTA+ complexes start to replace free CTA+ surfactants at the interface and thus reduce the interfacial electric field in the process. This causes the O-H bands from interfacial H2O molecules in our SFG spectra to decrease substantially, which reach a local minimum in intensity close to equimolar concentrations. Once electrostatic repulsion is fully screened at the interface, hydrophobic PSS-/CTA+ complexes dominate and tend to aggregate at the interface and in the bulk solution. As a consequence, adsorbate layers with the highest film thickness, surface pressure, and dilatational elasticity are formed. These surface layers provide much higher stabilities and foamabilities of polyhedral macroscopic foams. Mixtures around this concentration show precipitation after a few days, while their surfaces to air are in a local equilibrium state. Concentrations >0.1 mM result in a significant decrease in surface pressure and a complete loss in foamability. However, SFG and surface dilatational rheology provide strong evidence for the existence of PSS-/CTA+ complexes at the interface. At polyelectrolyte concentrations >10 mM, air-water interfaces are dominated by an excess of free PSS- polyelectrolytes and small amounts of PSS-/CTA+ complexes which, however, provide higher foam stabilities compared to CTAB free foams. The foam structure undergoes a transition from wet to polyhedral foams during the collapse.

Introduction

Foams are of great importance in many applications such as lightweight construction and heat insulation as well as in food or cosmetic products.[1−4] Mixtures of oppositely charged polyelectrolytes and surfactants also play a major role in cosmetics, pharmaceuticals, detergents, and mineral processing. Because liquid–gas interfaces are a major hierarchical element in foams, it is essential to understand the relation between molecular building blocks at gas–liquid interfaces and macroscopic foam properties such as foamability and stability via structure–property relationships. The latter can be used to predict and to tailor foam properties in a targeted way.

Both coverage and structure formation of molecules at the interface are major factors that can dominate the stability of foam.[5] That is because the latter determine the molecular interactions at the interface and the energy demand to create gas–liquid interfaces, which are both driving forces that determine bubble coalescence, foam drainage and Ostwald ripening.[6,7]

Changing interactions between molecules at the interface from repulsive to attractive regions is therefore one possibility to change interfacial and thus foam properties. This can be achieved by suppressing repulsive electrostatic interactions in a way that other interactions such as hydrophobic or van der Waals interactions become dominant. A possible way to realize the latter are formulations with positively charged surfactants and negatively charged polyelectrolytes in a concentration range where a point of zero net charge is being crossed at the interface. Surface and bulk behavior of oppositely charged polyelectrolyte/surfactant mixtures have been previously discussed in the literature,[8−16] where it was shown that the polyelectrolyte/surfactant ratio can have a substantial influence on foam stability.[12,17] Formulations with mixtures of tetradecyltrimethylammonium bromide (MTAB) and poly(acrylamidomethylpropanesulfonate) sodium salt (PAMPS) showed the highest stability of foam films when an enhanced synergistic adsorption of both species at the interface is observed at monomer concentrations of 0.1 and 1 mM for MTAB and PAMPS, respectively.[9] Surface excess measurements revealed complex changes in the surface adsorption behavior when changing the PAMPS concentration, showing that there is a depletion of both molecules from the air–water interface when the PAMPS concentration is increased to 3 mM, which also corresponds to a lower foam film stability.[9] Tensiometry of PAMPS/DTAB (dodecyltrimethylammonium bromide) mixtures showed lower surface tensions when coadsorption of the weakly surface active polyanion and the cationic surfactants took place and a highly surface active complex at the interface was proposed while binding in the bulk was negligible.[18] Screening of net charges in mixed layers of polyelectrolyte and surfactant is shown to reduce the stability of foam films and is caused by a reduction of electrostatic repulsion between the surfaces[19,20] that stabilize foam films. However, the stability of macroscopic foams is shown to have a maximum when net charges are screened.[21] Here, the major dynamic effects that are crucial for foam stabilization are Ostwald ripening, foam drainage, and bubble coalescence.[7] Bubble coalescence and Ostwald ripening are inherently controlled by the properties of the ubiquitous air–water interfaces, while drainage is effected by the bubble size distribution and the bulk viscoelasticity.[22−24]

Nevertheless, the exact nature of the adsorption mechanism and the influence of the charging in the system are yet not fully understood. Petkova et al.[17] showed that strongly interacting systems like cationic surfactant and anionic polyelectrolytes provide enhanced foam stability, but reduced foam capacity. However, clear synergistic effects on surface tension lowering and/or foam capacity were observed independent of the nature of surfactants and polyelectrolytes (anionic, cationic, nonionic).[17,25,26] Consequently, the stability of macroscopic foams from mixtures of polyelectrolytes and surfactants can be enhanced not only by charge screening but also by other effects that demand further studies.

In order to understand the properties and the composition of the interface, it is equally important to understand the phase behavior of the bulk system. Meszaros et al.[27] and later Abraham et al.[8,28] have shown that oppositely charged polyelectrolyte/surfactant systems might be far from an equilibrium state even if the surface appears to be in a (local) equilibrium. Investigations of such mixtures in the bulk have provided evidence that in concentration regions around the point of zero net charge, the oppositely charged components and their complexes tend to form aggregates and phase separation in the bulk is taking place due to a lack of colloidal stability. Resolving interfacial properties is important, however, the latter also shows that it is equally important to address possible aggregate formation and precipitation in the bulk solution over long time periods. The effects of the colloidal stability of oppositely charged polyelectrolyte/surfactant mixtures on the surface properties have been widely discussed in the literature. Here, major contributions come from Campbell and co-workers.[29−33]

In our work, we report on a unique combination of interface specific techniques such as ellipsometry, SFG spectroscopy, and tensiometry as well as macroscopic foaming experiments to investigate the structure–property relationship between surfactant/polyelectrolyte modified air–water interfaces and macroscopic foam. Vibrational sum-frequency generation (SFG) spectroscopy has been previously applied to study interfaces of polymers,[34] surfactant,[35] and proteins,[36] however, only a few studies exist that address polyelectrolyte/surfactant mixtures[37,38] and none of these show specific bands of each component in different spectral regions. By addressing both solvent and adsorbate specific bands, we gain detailed information on the molecular composition, charging state, and structure of the interfacial layer.

Materials and Methods

Sample Preparation

Hexadecyltrimethylammonium bromide (CTA+/Br–, CTAB; purity >99%) was purchased from G-Biosciences and was used as received. Deuterated CTAB (purity >99%, degree of deuteration >99.2%) was purchased from CDN Isotopes and used as received. Poly(sodium 4-styrenesulfonate) (Na+/PSS–, NaPSS) with an average molecular weight of about 70 000 g/mol was purchased from Sigma-Aldrich and was used as received. Stock solutions were prepared by dissolving the necessary amounts of surfactant and polyelectrolyte powders in ultrapure water (18.2 MΩcm; total oxidizable carbon <5 ppb). Subsequently, the stock solutions were sonicated for at least 10 min until full dissolution was reached. The mixtures were prepared by adding the necessary amounts of CTAB stock solution, NaPSS stock solution and water (in this order). As the formation of kinetically trapped aggregates can be affected by local concentration gradients that occur during the mixing of the solutions, the mixing protocol could influence the surfactant to polyelectrolyte binding ratio.[8,32,33] For that reason, also a different mixing protocol has been employed for the measurements of electrophoretic mobility and optical density. Here, we mixed identical volumes of stock solutions with twice the concentration as required in the final concentration. However, the mixing protocol seemed to have negligible effects on the results (see Supporting Information Figure S7). The CTAB concentration was fixed to 0.1 mM while the NaPSS concentration was varied from 10 μM to 20 mM. Note that NaPSS concentrations refer to the respective monomer concentration. Mixtures of NaPSS and CTAB had a pH of 6.4 ± 0.2 which was found to be independent of the NaPSS concentration.

Prior to the experiments the necessary glassware was stored in a bath of concentrated sulfuric acid (98% p.a., Carl Roth) with NOCHROMIX (Sigma-Aldrich) for at least 12 h and thoroughly rinsed with ultrapure water, subsequently. All experiments were performed at 297 K room temperature.

ζ-Potential Measurements

ζ-potentials of the CTA+/PSS– mixtures were determined with a Malvern Instruments Zetasizer Nano ZS. For each concentration, three measurements were performed and the mean value was calculated.

Turbidity Measurements

Turbidity measurements were done by measuring the transmission with a UV–vis spectrophotometer (Cary 100 Scan, Varian) at 25 °C. Samples were measured in plastic cuvettes of 10 mm path length and the optical density was evaluated at a wavelength of 450 nm. Measurements were done immediately after sample preparation and in consecutive time intervals. After a few days, the transmission of the supernatant was measured for those samples where precipitation took place.

Ellipsometry

Thicknesses of CTA+/PSS– adsorbate layers at the air/water interface were measured with a phase modulated Picometer ellipsometer (Beaglehole Instruments, New Zealand) operated at a wavelength of 632.8 nm. The sample was poured into a Teflon Petri dish with a diameter of 10 cm. For each sample, a time-resolved measurement was performed prior to an angle scan to ensure that the interfaces were equilibrated. Subsequently, angle scans between 51° and 55° versus the surface normal were performed with a step width of 0.5° at five different positions and with three measurements per position. The data were fitted using a three-layer model with refractive indices of 1.33 for the electrolyte subphase, 1.40 for the adsorbate layer, and 1.00 for the gas phase. The refractive index of the polyelectrolyte/surfactant layer was set to 1.40; as it is not a priori known, this assumption can lead to systematic errors that we will discuss below. However, independent of the choice for the refractive index we can compare the measured thicknesses on a qualitative level, for example, on an arbitrary length scale to identify trends.

Tensiometry

The surface pressure of surfactant/polyelectrolyte mixtures was determined via drop shape analysis of a pendant drop in a Krüss Dynamic Surface Analyzer (DSA 100S). Drops of equal volume were generated with a syringe and a video of the changes in drop shape as a function of surface age was recorded for at least 30 min. The video was analyzed and the surface tension after 30 min was calculated using the Young–Laplace equation.[39,40] The surface pressure Π(t) = γ0 – γt was then calculated from the surface tension with γ0 = 72.8 mN/m and γt being the surface tension of the neat air–water interface and of the CTA+/PSS– modified air–water interface after the adsorption time t, respectively.

In order to determine the viscoelastic properties, each drop was subjected to sinoidal oscillations with a constant frequency of 0.1 Hz and a mean change in relative drop surface area of ΔA/A = 0.07 for 100 s. After applying a Fourier transformation to the change of interfacial tension, we obtain the storage modulus E′ and the loss modulus E″.[39−41]

Each of the above procedures was repeated at least three times and the results were averaged.

Broadband Sum-Frequency Generation (SFG) spectroscopy

Broadband SFG spectroscopy is a nonlinear optical technique where a tunable broadband infrared pulse with a frequency of ωIR and a narrowband visible pulse with a fixed frequency ωvis are overlapped spatially and temporally at the surface of interest to generate photons with the sum frequency (SF) ωSF = ωIR + ωvis of the two laser beams. Our SFG measurements were performed with a home build broadband SFG spectrometer that is described elsewhere.[36,42] In this work, the IR and the visible beam were overlapped at the air–water interface after pouring 2.2 mL of a sample solution into a Petri dish. The reflected SF photons were collected and SFG spectra were recorded with an Andor Shamrock 303i spectrograph and an Andor iStar intensified charge-coupled device camera. Spectra in the range of C−H and O−H stretching vibrations (2780–3800 cm–1) were recorded by tuning the frequency of the IR beam in seven steps. The total acquisition time for a SFG spectrum in this region was between 4 and 16 min, depending on the respective signal intensity. Spectra in the range of S−O stretching vibrations (950–1300 cm–1) were obtained with three different IR frequencies in this range and total acquisition times between 2 and 8 min. All spectra were recorded with s-polarized sum-frequency, s-polarized visible, and p-polarized IR beams and were normalized to a reference spectrum of an air–plasma cleaned polycrystalline Au film.

For materials with centrosymmetry, SFG is surface specific and the intensity of the sum-frequency beam is proportional to the square of the second-order nonlinear optical susceptibility χ(2) of the material.[43,44] For centrosymmetric materials, all tensor components of χ(2) are zero in the bulk material that therefore does not contribute to the SFG signals in dipole approximation. Interfaces of centrosymmetric materials such as liquids and gases necessarily break the bulk centrosymmetry and thus give rise to dipole-allowed SFG signals. The SFG intensity can be expressed as a function of the intensities of the impinging laser beams IIR and Ivis as well as by a resonant and a nonresonant second-order nonlinear susceptibility χNR(2) and χ(2), respectively.[34,44]

The resonant part χ(2) depends on the resonance frequency ω and the bandwidth Γ of the vibrational mode k. The amplitude A is additionally a function of net molecular orientations at an interface. Assuming that A is the oscillator strength of a molecule with a hyper-polarizability β and an interfacial number density N, A can be expressed by an integral over all possible molecular orientations at the interface:

Consequently, the polarity of A (whether it is >0 or <0) and thus the SFG intensity and the shape of SFG spectra depend strongly on the orientation of interfacial molecules that give rise to a particular vibrational band.[34,36,38,44−46]

Foam Characterization

Foams were produced and characterized with a dynamic foam analyzer device DFA 100 (Krüss, Germany). The device allows determining the foam height as a function of foam age by measuring the light transmission through the foam column. For that, blue light from an LED panel is directed through the foam column and is detected by a line sensor at the opposite side.

Foams were produced by pouring 40 mL of a sample solution into the glass column with a length of 25 cm and a diameter of 4 cm. Subsequently, ambient air with a flow rate of 5 mL/s was pressed for 30 s through a porous glass frit (Carl Roth, Germany) that had pore sizes of 16–40 μm and which was positioned at the bottom of the glass column. Foam stabilities (FS) in percent are defined by FS = Vt/V0 × 100, where Vt and V0 are the respective foam volumes at a certain age t and at t = 0 s where the gas flow that produces the foam was stopped. Foam capacities (FC) in % are defined by FC = (V0/Vs) × 100 with the initial foam volume V0 at the point where the gas flow was stopped and the volume of the solution Vs. At least three foams per concentration were produced and the results were averaged. Additionally, the foams were investigated in terms of foam structure during aging. For that, a structure module (Krüss, Germany) consisting of a glass column with a prism, an LED, and a camera was used. Using this setup, it is possible to analyze a cross section of up to 285 mm2 in terms of structure and bubble size distribution.

Experimental Results

Determination of the Bulk ζ-Potential

ζ-potentials of NaPSS/CTAB mixtures were measured as a function of the polyelectrolyte concentration while the surfactant concentration was fixed to 0.1 mM. From a close inspection of Figure a, it becomes obvious that there is a clear transition from positive to negative ζ-potentials at NaPSS concentrations between ∼0.08 and ∼0.09 mM. These changes are attributed to a complex formation between PSS– and CTA+ ions due to electrostatic attraction, which causes positive net charges at excess CTA+ and negative net charges at excess PSS– concentrations. At low PSS– concentrations, this behavior points to overcharging of polyelectrolyte/surfactant complexes, while high PSS– concentrations result into negatively charged polyelectrolyte complexes with CTAB. Because all concentrations refer to the number of monomers in the PSS– polyelectrolyte, we propose that there is quite efficient binding of the surfactant to the polyelectrolyte: At 0.08 to 0.09 mM concentrations of PSS–, charge neutralization takes place, and at 0.1 mM bulk CTAB there is ∼0.01 to ∼0.02 mM free surfactant. When we compare this behavior to the previous work by Abraham et al.[8] who investigated the electrophoretic mobility of NaPSS/DTAB complexes and found a charge reversal at a surfactant/polyelectrolyte ratio of about 10, the driving force for CTA+/PSS– binding is much higher as compared to the previously studied DTAB surfactants. The origin for this effect could be related to the chain length of the applied surfactants.

Figure 1

(a) Bulk ζ-potential and (b) optical density at 450 nm as a function of bulk NaPSS concentration at a fixed bulk CTAB concentration of 0.1 mM, measured directly after mixing (black square) and after 4 days (red circle). The blue point (blue circle) indicates optical density of 20 mM NaPSS solution without CTAB. Solid lines are a guide to the eye. The vertical dashed line indicates concentration at point of zero charge in the bulk.

For the following results and their discussion, near equimolar concentrations are of great interest because a zero net charge can be a driving force to assemble adsorbate layers with mainly attractive intermolecular interactions at air–water interfaces.[47,48] It is likely that CTA+/PSS– complexes have a higher hydrophobicity than the individual molecules that can consequently lead to an enhanced interfacial adsorption of such complexes. In addition, these soft aggregates might already form in the bulk solution and can have a stabilizing effect on the macroscopic foams similar to well-known Pickering foams that have been stabilized by hard colloidal particles.[49]

Results from Turbidity Measurements

From a close inspection of Figure b, it is obvious that aggregates of hydrophobic CTA+/PSS– complexes are being formed in the bulk at concentrations lower than 0.09 mM and are the cause for the observed sample’s turbidity. This effect is maximal at 0.09 mM concentration and is within the experimental scatter, which is consistent with the bulk point of zero net charge at ∼0.08 mM (Figure a). Obviously, there exists a lack of colloidal stability at concentrations around 0.08 mM that is caused by the absence of strong electrostatic repulsion and thus determines the critical aggregation concentration of the present system. For higher concentrations, the turbidity of the sample is zero because there are sufficient negative charges to stabilize the polyelectrolyte/surfactant complexes and free polyelectrolytes by electrostatic repulsions. Going to even higher concentrations (>1 mM), the optical density increases again but is caused by the high concentration of polyelectrolyte and not by any interactions with CTAB because UV–vis spectra of NaPSS solutions with and without CTAB are identical in this concentration region (Figure b). After 2 days, flocculation is observed for concentrations >0.04 and <0.12 mM NaPSS. Consequently, the optical density of the supernatant is decreasing remarkably.

Results from Ellipsometry

Assuming a constant refractive index (which will be challenged below) for CTA+/PSS– adsorbate layers at the air–water interface, relative changes in layer thickness were determined for NaPSS concentrations of 0.01 to 20 mM and are presented in Figure a. For low NaPSS concentrations, layer thicknesses of ∼1 nm were determined that are representative for layers dominated by CTA+ ions with low molecular order and in a close to flat lying configuration at the interface. Note that our vibrational SFG spectra bring major support to this conclusion (see below).

Figure 2

(a) Thickness of adsorbate layers from ellipsometry, (b) surface pressure, and (c) surface dilatational storage modulus of CTA+/PSS– modified air–water interfaces as a function of bulk NaPSS concentration and at a fixed bulk CTAB concentration of 0.1 mM. (d) Foam stability (FS) for foam ages of 60, 300, and 900 s. Red data points (red circle) indicate values of 20 mM NaPSS solution without CTAB. All solid lines are a guide to the eye. The vertical dashed line indicates concentration at point of zero net charge in the bulk.

Concentrations between 0.02 and 0.05 mM NaPSS with CTAB/NaPSS molar ratios between 2 to 5, resulted in layer thicknesses of ∼2 nm, which is in excellent agreement with layer thicknesses reported in a previous neutron reflectometry study by Taylor et al.[50]

In their study Taylor et al. have used CTAB/NaPSS molar ratio of about 3.5 and report a layer thickness of 2.0 nm. Consequently, the earlier results from Taylor et al. bring strong support to our ad hoc assumption for the refractive index of the surface adsorbed molecules at these concentrations.

A further increase in NaPSS concentration leads to much higher layer thicknesses and a pronounced local maximum of ∼3.8 nm at 0.09 mM NaPSS. This indicates a close relation between the thickness of adsorbate layers and the charging behavior of molecules in the bulk (panel a). For concentrations >0.1 mM, the layer thickness decreases again and reaches a local minimum of 0.6 nm at 0.2 mM whereas higher concentrations lead to a second increase in layer thickness to 3.6 nm at 20 mM NaPSS. Because 20 mM NaPSS solutions without CTAB were also measured and showed much thinner films with thicknesses of ∼2 nm, we can conclude that there is a significant effect of CTAB on the molecular structure of the interfaces also at the highest NaPSS concentrations (>0.2 and ≤20 mM). At this point, it should be noted that we have used for our analysis a constant refractive index for the adsorbate layer that was not allowed to change as a function of NaPSS concentration. At this early stage of our discussion, this is an instrumental approach. However, in light of our results from vibrational SFG (below) the above assumption needs to be modified and will lead to a much better understanding of the wetting behavior of CTA+/PSS– complexes at the air–water interface.

Results from Tensiometry

Figure b presents the results for the surface pressure after a time of 30 min. The changes in surface pressure follow the same trend as we have already discussed for the layer thickness (Figure a), where the local maximum in layer thickness at 0.09 mM NaPSS roughly corresponds to a local maximum in surface pressure Π of ∼23 mN/m at 0.08 mM. In addition, the local minimum in layer thickness between 0.5 and 1 mM NaPSS is accompanied by negligible surface pressures. For concentrations between 0.5 and 1 mM, we observed surface tensions (negligible surface pressures) that were similar to those of a neat water surface and may be interpreted in a way that surface excess is low. In contrast to the substantial decrease in surface pressure, the decrease in SFG intensities of PSS– specific bands is much less extensive (see below). Similar changes were reported for mixtures of PAMPS polyelectrolytes and MTAB surfactants in a neutron reflectometry study by Fauser et al.[9] where an increase of surface tension from 52 to 67 mN/m but layer thicknesses between 1 to 2 nm were found for all samples.

In addition to the surface pressure, we have also investigated the surface dilatational storage E′ and loss modulus E″ of the adsorbate layer after the interface was in an equilibrated state (Figure c). The storage modulus E′ increases from 12 mN/m at concentrations of <0.03 mM NaPSS to a local maximum of 89 mN/m at 0.09 mM NaPSS. For concentrations of 0.1 to 10 mM, E′ decreases and is between 38 and 57 mN/m. For concentrations >10 mM, a second increase in E′ is observed with values of 100 mN/m at 20 mM. The loss modulus E″ shows a similar behavior but is considerably lower with values <10 mN/m (Supporting Information).

The stability of macroscopic foam can be described by three distinct NaPSS concentration regions (Figure d). For concentrations <0.08 mM, foams can be produced with rather constant foam capacities of ∼345% (Supporting Information). A maximum in foam stability is observed at ∼0.09 mM. A slight increase in foam capacity between 0.01 and 0.1 mM from ∼340 to ∼440% shows good correlation with the previously discussed increase in surface pressure in this concentration region. For concentrations between 0.1 and 3 mM, we find negligible foamabilities, which seem to be linked to the negligible surface pressures in this concentration region. With even higher NaPSS concentrations, foams are produced again and both foam capacity and stability reach values similar to those of the first local maximum near equimolar concentrations. As expected from the surface pressure and layer thickness, the presence of CTAB has also a promoting effect on the foam stability that can be seen from the foam stability for a 20 mM NaPSS solution without CTAB.

Foam structure analysis as shown in Figure indicates wet foams with rather small bubbles for an excess of CTAB (0.01 mM NaPSS), while polyhedral foams are produced from equimolar mixtures. A transition from wet foams to polyhedral foams is also observed but at NaPSS concentrations >10 mM. The foams generated by an excess of NaPSS show the smallest bubble sizes after foaming, but during the collapse a rapid increase in bubble size caused by bubble coalescence and drainage is found that is likely the origin of the structural transition from wet to polyhedral foams.

Figure 3

Foam structure directly after (t = 0 s) initial foaming for 30 s and during aging of foams for different bulk NaPSS concentrations. Aging times are as indicated. Columns indicate liquid (dark blue) and foam (light blue) heights. The scale bar is for all images as shown in the left-hand side of the third row.

Results from Vibrational SFG Spectroscopy

SFG spectra were recorded for bulk NaPSS concentrations of 0.01 to 20 mM with 0.1 mM CTAB. Choosing this range, it is possible to observe effects that are predetermined from the molecular properties in the bulk where we observe a reversal in net charge of CTA+/PSS– complexes at ∼0.09 mM (Figure a). In Figure , we present SFG spectra for two different frequency regions, 950–1300 and 2780–3800 cm–1. Three distinct vibrational bands are visible in the frequency region of 950–1300 cm–1 and are indicative for the presence of PSS– molecules and their complexes with CTA+ ions at the air–water interface. In particular, vibrational bands centered at 1011 and 1130 cm–1 are attributable to in-plane bending and in-plane skeleton vibrations of PSS– aromatic moieties. In addition, the third vibrational band at 1042 cm–1 can be assigned to symmetric stretching vibrations of sulfonate groups at the interface.[51]

Figure 4

Vibrational SFG spectra of PSS–/CTA+ complexes at the air–water interface for (a) S—O and aromatic stretching regions, (b) C—H (2800–3070 cm–1) and O—H (3000–3800 cm–1) stretching regions, and (c) a close up of the stretching band at 3700 cm–1 due to dangling O—H. Bulk CTAB concentration was 0.1 mM for all samples, andNaPSS concentrations were as indicated on the right.

In the frequency region of 2800–3000 cm–1, we observe vibrational bands that originate from R−CH2 and R−CH3 stretching vibrations (Table ) and another band centered at 3060 cm–1 which is caused from aromatic C−H stretching vibrations of interfacial PSS–. Additionally, broad vibrational bands centered at 3200 and 3450 cm–1 are due to symmetric O−H stretching vibrations of hydrogen-bonded interfacial water molecules.[36,44,47]

Table 1

Assignments of Vibrational Bands in SFG Spectra (Figure ) of CTA+/PSS– Complexes Adsorbed to the Air–Water Interfacea

band	[cm–1]	ref	band	[cm–1]	ref
arom. CH	1011	(51)	CH3 (F)	2936	(36, 47)
R-SO3– (ss)	1042	(51)	arom. CH	3060	(36)
arom. CH	1130	(51)	OH (ss)	3200	(36, 44, 47)
CH2 (ss)	2850	(47)	OH (ss)	3450	(36, 44, 47)
CH3 (ss)	2875	(36, 47)	OH (free)	3710	(36, 44)

(F) and (ss) stand for Fermi resonance and symmetric stretching vibrations, respectively.

Table presents an overview of all vibrational bands in SFG spectra of PSS–/CTA+ modified air–water interfaces and their band assignments.

Looking at the spectral region from 970 to 1230 cm–1 (Figure ), it becomes obvious that higher concentrations of NaPSS lead to a dramatic increase in SFG intensities of all vibrational bands in this frequency region that is indicative for an increase of the PSS– surface coverage. In order to provide a more quantitative analysis, we have fitted the SFG spectra in Figure a with model functions according to eq where we have allowed the amplitude as well as the frequency of all bands as free parameters in our fitting procedure with Lorentzian line shapes. The SFG amplitudes of all three bands are shown in Figure a. An increase of SFG amplitudes can be seen up to 0.1 mM NaPSS, where a local maximum exists. For concentrations of 0.2 to 10 mM, the amplitudes are slightly lower as for 0.1 mM but nearly independent of the NaPSS concentration. For 20 mM NaPSS, a further increase of amplitudes is visible.

Figure 5

(a) SFG amplitudes of the vibrational bands at 1011, 1042, and 1130 cm–1; (b) SFG intensity of O−H stretching bands from interfacial water molecules. Solid lines are a guide to the eye. The vertical dashed line indicates concentration at the point of zero net charge in the bulk (see Figure a).

Analysis of the O−H intensities around 3200 and 3450 cm–1 (Figure b) provides information on the charging state of the interface and on both orientation and polarization of water molecules. For an excess of CTAB in the bulk (NaPSS < 0.08 mM), there is likely to be also an excess of positive charges at the air–water interface, which is the origin of the high intensity of the broad O−H bands around 3200 and 3450 cm–1 from hydrogen-bonded interfacial H2O. As the NaPSS concentration is increased, the O−H intensities decrease dramatically to a minimum with nearly zero intensity at 0.08 to 0.1 mM concentrations. For higher concentrations, there is a sharp increase until similar apparent intensities are reached as for interfaces dominated by free CTA+ (Figures and 5b). We attribute these changes to electric field-induced effects caused by adsorption of CTA+/PSS– complexes with different net charges. These effects are similar to what is observed for protein/surfactant mixtures at the air–water interface.[47] The origin of these changes is the coherent process of sum-frequency generation where the oscillator strength A of a SFG active band is not only a function of the number density N of adsorbed molecules but also a function of their orientation and polarization (eq ). This means that for a charged interface, strength, and polarity of A from interfacial H2O molecules are closely connected to the strength and orientation of the mean electric field at the interface. Changes in polarity of A from O−H vibrational bands can be seen by a close inspection of Figure b, where the interference between O−H stretching and the aromatic C−H stretching band at 3060 cm–1 leads first to a positive going band at low NaPSS concentrations and a negative going and highly dispersive band at high NaPSS concentrations. This change in polarity of the aromatic C−H band is thus caused by changes in the net orientation and polarization of interfacial water molecules. At low NaPSS concentration, a positive net charge does exist at the interface and originates from an excess of free CTA+ ions, whereas at high NaPSS concentrations PSS– polyions are the charge determining surface species (Figure a) and, thus, lead to surfaces with a negative net charge. Accordingly, the change in polarity at concentrations where the O−H intensities are very small (∼0.09 mM NaPSS) can be attributed to a charge reversal at the interface where the interfacial point of zero net charge is being crossed.

Figure c presents SFG spectra in the frequency region around 3700 cm–1 in more detail. At nearly equimolar concentrations of 0.1 mM there is a clear rise of a weak but noticeable band at 3700 cm–1. At concentrations above and below 0.1 mM, the broad O−H bands from hydrogen bonded interfacial H2O dominate the spectra. Whether the band at 3700 cm–1 is present at these concentrations or not cannot be determined because it cannot be distinguished from the dominating bands from H-bonded H2O. For neat air–water interfaces without surface adsorbed molecules, the stretching vibration of free O−H groups is observed (Supporting Information). This “free O−H” band is attributable to water molecules that are not fully hydrogen bonded and have free O—H bonds sticking into the gas phase.[36,44] Shen and co-workers[52] have estimated that for a neat water surface roughly 20% of the interfacial H2O molecules have free O−H groups while the remaining molecules are fully hydrogen bonded. Usually even small amounts of a surface active species cause the free O−H band to decrease dramatically in intensity.[36,46,53,54] As a consequence, the occurrence of a free O−H band is a clear sign that the interface is not fully occupied.

Discussion

Combining the results from ellipsometry and SFG spectroscopy as well as bulk ζ-potential, turbidity, and surface tension measurements, we can now deduce a more complete molecular picture of the building blocks at the air–water interface that drive properties of macroscopic foam such as stability, foamability, and structure.

For that purpose, we focus the discussion on four concentration regions that reflect the net charge at the interface and cause different behavior of macroscopic foam. The first region (i) is represented by concentrations ≪0.1 mM NaPSS, which are below the point of zero net charge in the bulk and at the interface. The second region (ii) is around the point of zero net charge whereas we define a third (iii) region to concentrations between 0.1 and 10 mM (no foaming and negligible surface pressure). In a fourth region (iv) at concentrations >10 mM NaPSS, foaming ability has recovered while the surface pressure has also increased again.

Region 0.1 mM CTAB with ≪0.1 mM NaPSS Concentrations

In this region, air–water interfaces are dominated by free CTA+ ions at very low polyelectrolyte concentrations. This is corroborated by the bulk ζ-potential of +13 mV at 0.01 mM NaPSS (Figure a) that is caused by overcharging of PSS– polyelectrolytes with excess CTA+. In addition, the thickness of the adsorption layer is close to 1 nm (Figure a) and corresponds to the thickness of a CTA+ dominated interface with low molecular order. The latter can be deduced from the substantial portion of CH2 symmetric stretching vibrations (2850 cm–1) to the SFG spectra in Figure b, which is in the case of a CTA+ dominated interface indicative of a high density of Gauche defects and/or flat lying molecules. In the opposite case with an ordered network of CTA+ molecules that have their alkyl chains in an all-trans configuration, CH2 contributions to SFG spectra necessarily need to be negligible because of local centrosymmetry that leads to a cancellation of such signals (eq ).[42] This is obviously not the case and as a consequence it is not surprising that at these concentrations wet but relatively instable foams can be generated with great structural similarities as seen for foams from pure CTAB solutions (Figure ).

At higher NaPSS concentrations, the thickness of the adsorbate layer increases and is accompanied by a substantial decrease in O−H stretching intensities from interfacial H2O (Figure b). Such a decrease in intensity is consistent with a preferential adsorption of CTA+/PSS– complexes and their aggregates that have lower net charges but higher hydrophobicities. Further support to this conclusion comes from the notable increase in surface pressure (Figure b) and from a strong rise in SFG amplitudes of vibrational bands that are directly associated with PSS– molecular groups (Figures a and 5b). From electrostatic considerations, it is likely that CTA+/PSS– complexes and larger aggregates are already formed in the bulk, which is also seen in our ζ-potential and turbidity measurements (Figure ). Besides electrostatic interactions, additional hydrophobic interactions of the complex due to the presence of the CTA+ aliphatic chains and the aromatic ring residues of the polyelectrolyte can occur and may help to stabilize CTA+/PSS– complexes at air–water interfaces. Turbidity measurements show that aggregates are forming instantly after mixing the components for NaPSS concentrations from 0.05 to 0.12 mM. These bulk aggregates can enter the air–water interface, an assumption that is corroborated by change in dynamic surface tension at these concentrations (see Supporting Information). At the surface, aggregates form layers of around 2 nm thickness that stabilizes the gas bubbles in the macroscopic foam. SFG, tensiometry, and ellipsometry show that the interface seems to be in a local equilibrium on short time scales <1 h, while turbidity measurements show that the bulk solution is still in a nonequilibrium state on a much longer time scale of days. For concentrations <0.05 mM, the samples are stable in the bulk for at least 1 week, which is probably due to sufficiently large electrostatic repulsion that prevents the formation of (larger) aggregates. This points to adsorption of simple polyelectrolyte/CTAB complexes rather than to the presence of larger aggregates at the interface. For concentrations <0.02 mM, the absence of strong PSS– specific vibrational bands at 1042, 1130, and 3060 cm–1 (Figure , Table ) and the shape and intensity of SFG spectra in the C−H and O−H region (Figure b), which are nearly identical to those from pure 0.1 mM CTAB solutions (Supporting Information), show that there is too little PSS– in the bulk in order to cause a noticeable change to the surface properties. Accordingly, foams produced from solutions with these concentrations will be dominated by interfacial layers which are a mixture of aggregates from CTA+/PSS– complexes and free CTA+ ions.

Region 0.1 mM CTAB with ≈0.09 mM NaPSS Concentration

In the (ii) concentration region where a charge reversal is observed, the interfaces are dominated by CTA+/PSS– complexes that have attractive lateral interactions but negligible electrostatic repulsions. This is directly associated with the absence of strong O−H stretching bands (Figure b) and with a local maximum in surface elasticity E′ (Figure c, ∼0.1 mM NaPSS). The optical density of freshly prepared samples is the highest around 0.09 mM (Figure b). On the other hand, the prevailing hydrophobic interactions of the adsorbate with the aqueous phase lead to a more compact packing and a pronounced maximum in layer thickness of 3.8 nm (Figure a). This dense network of molecules and their aggregates with attractive lateral interactions and high coverage is obviously responsible for the local maximum in foam stability. This behavior is similar to what is observed for protein foams with canceled electrostatic interactions that was achieved by either choosing a pH close to the bulk isoelectric point of the protein or by a certain mixing ratio of proteins with oppositely charged surfactants.[47,55] Similar tuning of foam properties has also been found by Argillier et al.,[56] who showed that additions of small amounts of polyelectrolytes to an oppositely charged surfactant caused the stability of foams and emulsions to increase. Another possible stabilization mechanism originates from a Pickering effect of the polyelectrolyte/surfactant aggregates that are forming in the solution directly after mixing similar to particle stabilized foams or emulsions.[57,58] Aggregates that form due to the absence of electrostatic repulsion have been also shown to enhance stability protein protein foams.[5,36,47]

Region 0.1 mM CTAB with 0.1 mM > NaPSS < 10 mM Concentrations

In the (iii) region, an abrupt drop of foam stability is accompanied by negligible surface pressures. This sudden change in surface tension for oppositely charged polyelectrolyte/surfactant mixtures is discussed in the literature.[9,29] While it is proposed that slow precipitation processes in the phase separation regime lead to depletion of surface active material from the bulk, this process is not completed in our case and the increase in surface tension is likely due to an increase in hydrophobicity of the complexes as described previously by Fauser et al.[9] The changes in surface pressure (Figure a) are much more substantial as compared to the changes in SFG intensities from polyelectrolyte specific bands (Figures and 5). The former can be taken as evidence for an apparent reduction in surface excess of polyelectrolyte/surfactant complexes. In fact, negligible surface pressures are observed but not all species are repelled from the air–water interfaces because CTAB (as evidenced by our experiments with deuterated molecules; Supporting Information) as well as PSS– specific bands are still observable with relatively high SFG intensities. Here, only a weak drop of the latter is observed in this concentration region.

This apparent contradiction can be resolved by the following hypothesis: At concentrations >0.1 mM, excess free PSS– molecules can be expected in the bulk solution and CTA+ ions can easily find free PSS– polyions as binding partners without compromising the latter colloidal stability. Some of these complexes can enter the air–water interface but do not lead to a substantial increase in surface pressure. However, in order to give rise to the strong SFG bands at low coverage the conformational order of these complexes at the air–water interface must be much higher as compared to the molecular structures at equimolar concentrations (compare eqs and 2). Some support for this hypothesis, comes from the broad but noticeable free O−H stretching band at 3700 cm–1 in our SFG spectra (Figure c). This band is necessarily only visible when water molecules are present in the topmost surface layer, which is typically the case when the surface coverage is low. Further support may be gained from our earlier conclusion that the interface at concentrations ≤0.1 mM is dominated by aggregates and free CTA+ ions. In particular, the molecular structures assumed by these aggregates are likely to be disordered and thus the decrease in SFG intensity due to their decrease in surface coverage can be partially compensated by the gain in intensity due to molecules with higher conformational order but lower coverage.

In addition to the results from SFG spectroscopy, surface dilatational rheology provides further evidence that the air–water interface is at concentrations >0.1 mM different from a neat water surface because the adsorbate layer has a relatively high elasticity E′ of ∼50 mN/m. This value is close to E′ at concentrations where we observe stable foams (Figure c,d) with moderate surface excess.

Similar observations have been reported by Kristen et al.,[10] who attributed the high surface tension at CTA+/PSS– molar ratios >1 to hydrophobic interactions between the aliphatic chain and the polymer backbone and proposed that the charged groups of both compounds point toward the solution, which makes the complexes more hydrophilic. This might be also a reasonable explanation for the observed effect, although we cannot provide direct evidence for this hypothesis from our results.

So far we have not discussed the bulk structure of the polyelectrolyte. At low concentrations, the structure is likely to be close to an elongated chain, however, Fundin and Brown[59] found that the NaPSS polyelectrolyte chain (60 kDa) can be contracted and can adopt the curvature of CTAB micelles. Although, in our experiments and in those by Kristen et al.[10] CTAB concentrations were below the bulk critical micelle concentration (CMC), Kogej and Škerjanc showed that NaPSS can induce aggregation of CTAB into micelles at concentration well below the CMC.[60] Here, Kogej and Škerjanc propose that stable aggregates between the polyelectrolyte and the surfactant are formed via the inclusion of the polyelectrolyte into the surface of the micelle-like surfactant aggregates.[60] Additional information that is consistent with CTA+ ions at the interface being in a micelle-like configuration comes from our SFG spectra in Figure b, where a dramatic decrease in CH2 contributions and a decrease in CH2 (2850 cm–1)/CH3 (2875 cm–1) intensity ratio is observed. These changes point to an interfacial molecular structure with the alkyl chains of CTA+ in a more centrosymmetric configuration. As discussed above, this is not the case at low concentrations of NaPSS or even in the absence of NaPSS.

Zero foam stabilities that are present in this concentration region are obviously due to the high solvation degree of the complexes and are directly coupled to the negligible surface pressure (high surface tension), which makes the formation and stabilization of gas bubbles energetically unfavorable. For that reason no foam is being generated.

Region 0.1 mM CTAB with >10 mM NaPSS Concentrations

Here, the surface pressure has increased again and relatively stable foams are formed. However, comparing the foam structure and stability in Figure , a cooperative effect of CTA+ is still noticeable as foam stability is higher in the presence of CTAB. In addition, the SFG spectra of NaPSS mixtures with deuterated CTAB molecules show negligible intensity from C−H bands but relative strong C−D stretching bands from interfacial (deuterated) CTA+ ions (see Supporting Information). This observation brings strong support to our conclusion that C−H bands at high NaPSS concentrations in Figure originate mainly from interfacial CTA+.

Conclusions

Using a unique combination of interfacial sensitive vibrational SFG spectroscopy and complementary analytical methods, we have identified molecular building blocks at air–water interfaces and provide new information on the adsorption behavior of the polyelectrolyte/surfactant complexes. This has allowed us to determine structure–property relations between interfacial building blocks and macroscopic foam stability and structure.

CTA+ dominated interfaces yield low foam stabilities. CTA+/PSS– complexes and aggregate with negligible net charges form at close to equimolar concentrations (0.09 mM) and give rise to thick adsorbate layers with attractive lateral interactions, high surface pressures, and dilatational elasticities E′. Although, local equillibrium states at the surface can be reached relatively fast, aggregation and phase separation in the bulk solution occurs on a days’ time scale. These aggregates substantially increase the foam stability but are largely absent at higher concentrations >0.1 mM. Here, air–water interfaces are covered by CTA+/PSS– complexes with high conformational order but low coverage, which leads to a dramatic loss in surface pressure and foam stability.