Ion Pairing and Adsorption of Azo Dye/C16TAB Surfactants at the Air-Water Interface.

Mixed layers of 6-hydroxy-5-[(4-sulfophenyl)azo]-2-naphthalenesulfonate (Sunset Yellow, SSY) and cetyltrimethylammonium bromide (C16TAB) at the air-water interface were studied using vibrational sum-frequency generation (SFG) and dynamic surface tension measurements. In the bulk, addition of C16TAB to SSY aqueous solution causes substantial changes in UV/vis absorption spectra, which originate from strong electrostatic interactions between the anionic SSY azo dye with the cationic C16TAB surfactant. These interactions are a driving force for the formation of SSY/C16TAB ion pairs. The latter are found to be highly surface active while free SSY molecules show no surface activity. Dynamic SFG as well as surface tension measurements at low SSY concentrations reveal that free C16TAB surfactants adsorb at the air-water interface on time scales <1 s where they initially form the dominating surface species, but on longer time scales free C16TAB is exchanged by SSY/C16TAB ion pairs. This causes a dramatic reduction of the surface tension to 35 mN/m but also in foam stability. These changes are accompanied by a substantial loss in SFG intensity from O-H stretching bands around 3200 and 3450 cm-1, which we relate to a decrease in surface charging due to adsorption of ion pairs with no or negligible net charges. For SSY/C16TAB molar ratios >0.5, the O-H bands in SFG spectra are reduced to very low intensities and are indicative to electrically neutral SSY/C16TAB ion pairs. This conclusion is corroborated by an analysis of macroscopic foams, which become highly instable in the presence of neutral SSY/C16TAB ion pairs. From an analysis of SFG spectra of air-water interfaces, we show that the electrostatic repulsion forces inside the ubiquitous foam films are reduced and thus remove the major stabilization mechanism within macroscopic foam.

Introduction

Azo dyes are widely used in many applications such as coloring additives in food products.[1] Every year, >105 tons of dye are produced of which roughly 10% end up in wastewater, where they can cause serious environmental problems.[2] As a consequence, the removal of azo dyes from aqueous solutions is mandatory in terms of environmental protection. However, the chemistry of dye treatment in wastewater is a challenging task, as azo dyes are very stable against degradation using light, heat, oxidizing agents, etc.[2,3] Common methods for dye removal are relying on approaches such as coagulation, flocculation, absorptive bubble separation, and foam fractionation.[4] Obviously, the interfacial properties of azo dyes are of great importance and in case the dye is not surface active itself, additives are used in order to transfer the dye to the air–water interface. For that reason, molecular interactions between surface active additives and azo dyes are an important research topic.[5−13]

To efficiently separate the azo dyes from wastewater by foam fractionation, it is necessary that the dye–surfactant complex not only has a high surface activity but also a high tendency to form stable foams. In order to tune foam properties like foamability and stability in a targeted way, the relation between the chemistry of liquid–gas interfaces, which are the dominant hierarchical element inside foam, and the properties of the macroscopic foam must be revealed. For that reason, the surface coverage, the charging state as well as the molecular structure and orientation of surface adsorbed species are of major importance for foam properties.[14,15]

In this work, we combine interface specific techniques with foaming experiments to gain insight into structure–property relations between macroscopic foam and the molecular structure of air–water interfaces. The latter interfaces are modified with mixtures of Sunset Yellow (SSY) (Figure ) and cationic cetyltrimethylammonium bromide (C16TAB) surfactants.

Figure 1

Chemical structure of the anionic dye Sunset Yellow (6-hydroxy-5-[(4-sulfophenyl)azo]-2-naphthalenesulfonate).

New information about the molecular structure and charging state of these interfaces in and outside equilibrium conditions is gained by the application of vibrational sum-frequency generation (SFG) and dynamic surface tension measurements. SFG has been shown to be a powerful tool for studies of adsorption processes at liquid interfaces,[16−20] but to the best of our knowledge little work has been done with SFG on the adsorption of azo dye–surfactant mixtures at air–liquid interfaces.[21,22] Here we expect that our results not only help to gain a better understanding of the interface chemistry, but also help to predict macroscopic foam properties.

Principles of Sum-Frequency Generation

SFG spectroscopy is an all optical technique for studying the molecular structure of interfaces with respect to their composition, surface coverage and molecular orientations.[23,24] For SFG, a visible laser beam with a fixed frequency at ωVIS and a tunable infrared laser beam at ωIR are overlapped spatially and temporally at an interface,[25,26] where they generate a third beam with the sum frequency (SF) of the frequencies coming from the two fundamentals eq .[27,28]

The SF intensity is proportional to the absolute square of the second-order nonlinear susceptibility χ(2). If centrosymmetry is given like in the bulk of liquids or gases, χ(2) is zero and, consequently, there is no symmetry allowed SF signal from the bulk. Interfaces, however, break the prevailing centrosymmetry of the bulk and SFG spectroscopy becomes inherently surface sensitive.[27,29] The second-order susceptibility χ(2) consists of a nonresonant part χ(2) and a resonant part χ(2). While the nonresonant contribution is mainly caused by electronic excitations, the resonant contribution originates from vibrational transitions when the frequency of the IR beam matches the frequency of a vibrational mode of molecules at the interface. The vibrational mode has to be both IR and Raman active.[24] In this case, the intensity of the SF beam is resonantly enhanced eq .[30,24]with the nonresonant andthe resonant parts of the second-order electric susceptibility. A ∝ N⟨α μ⟩, ω, Γ, and φ are the SF amplitude, frequency, bandwidth, and relative phase of the vibrational mode k. The SF amplitude depends on both the molecular number density N and the orientational average of the Raman polarizability αk and the dynamic dipole moment μk. If all interfacial molecules are perfectly polar ordered, the SF amplitude of a vibrational mode reaches a maximum. In contrast, a highly disordered interface or isotropic bulk molecules lead to low SF signals because the orientational average ⟨αμ⟩ and thus the SF amplitude equals zero.[23] Consequently, only a few molecular layers at the interface can contribute to the SF signal.

In the presence of an additional static electric field EDC, which can be induced by the net charge of surface adsorbed molecules, an additional third-order contribution χ(3) can occur and interfacial water molecules can become polar ordered and polarized near the charged interface and therefore additionally contribute to SF or second-harmonic signals.[30−32] In the past, this has often been used in order to get qualitative and in some cases even quantitative information about the charging state of aqueous interfaces.[15] With the wave vector mismatch Δk, the resulting nonlinear response is given by[33]

The strength of the static electric field at the interface E is mainly determined by the Debye length κ–1 and thus by the ionic strength as well as by the surface potential ψ0.[31−34] After integration, the following equation is obtained:[31−34]

Since χ(3) has the same frequency dependence as χ(2), the amplitude of a vibrational band in SFG spectra can be influenced by χ(3) effects and is thus dependent on the surface potential which is directly coupled to the surface charge σ0 by the Grahame equation.55

Materials and Methods

Sample Preparation

The required glassware was cleaned with Alconox detergent solution (Sigma-Aldrich) and stored in concentrated sulfuric acid (98% p.a., Carl Roth) with NOCHROMIX (Sigma-Aldrich) overnight. Afterward, it was thoroughly rinsed with ultrapure water until the acid was completely removed.

Hexadecyltrimethylammonium bromide (C16TAB; purity >99%) and Sunset Yellow (SSY; dye content 90%) were purchased from Sigma-Aldrich and used as received. Stock solutions were prepared by dissolving the necessary amount of surfactant and dye in ultrapure water (18.2 MΩ·cm; total oxidizable carbon <5 ppb). Afterward the stock solutions were sonicated until full dissolution was reached. The mixtures were prepared by adding the necessary amount of water, C16TAB stock solution (800 μM) and SSY stock solution (5 mM) (in this order). As azo dyes can show cis-trans-isomerization, the SSY stock solutions were stored and kept in the dark for 1 week before further treatment to avoid light induced changes of SSY. For the same reason all samples were prepared and measured under red light. The C16TAB concentration was fixed to 120 μM while the SSY concentration was varied from 0–120 μM. All experiments were performed at room temperature. In order to minimize the effect of possible sample aggregation (see Supporting Information) on the measuring results, all experiments as explained above were performed directly after sample preparation.

Tensiometry

The surface tension of Sunset Yellow/C16TAB mixtures was determined via drop shape analysis of a pendant drop in a dynamic surface analyzer, DSA100 (Krüss, Germany). Drops were generated at the end of a cannula with an outer diameter of 1.83 mm using a 3 mL syringe. To avoid drop shrinkage caused by evaporation, the drop is created in a water saturated atmosphere inside a custom designed vessel. The drop shape was recorded by a camera and the illumination was constrained to wavelengths >590 nm by using long-pass filters (Schott, RG 590) as vision panels implemented into the vessel to avoid light induced changes of the samples. A video of the changes in drop shape as a function of surface age was recorded until equilibrium was reached. Dynamic surface tensions were obtained by drop shape analysis using the Young–Laplace equation.[35,36] Each of the above procedures was repeated at least three times and the results were averaged.

Foam Characterization

A dynamic foam analyzer, DFA 100 (Krüss, Germany) was used to measure foam properties such as stability, capacity and foam structure. Foams were produced from 40 mL sample solutions which were poured into a glass column with 250 mm length and 40 mm diameter. A porous plate with pore sizes between 40 and 100 μm was fixed at the bottom of the column and ambient air was injected through the porous plate into the column and thereby homogeneously distributed over the cross section. The flow rate was set to 5 mL/s for all experiments and foams were produced by streaming air through the sample solution for 30 s. During foam formation and foam decay the foam height was determined as a function of foam age. For that, an infrared LED panel is installed at one side of the column and a line sensor at the opposite side of the column. The foam height is accessible by measuring the light transmission through the glass column. Foam stabilities (FS) were defined as FS = Vt/V0 × 100% and foam capacities (FC) as FC = (V0/Vs) × 100%, where Vt is the foam volume at foam age t, V0 the maximum foam volume. and Vs the volume of the sample solution (40 mL).

The DFA 100 device was also equipped with a foam structure module (Krüss, Germany) enabling foam structure analysis. For that, a prism is attached to the glass column which captures the foam lamellae at the wall of the glass cell. Images from the prism wall provide 2D cuts through the foam structure and were recorded as a function of foam age within a cross section of ∼285 mm2. Using the image analysis tool of the Krüss DFA 100 device, the mean bubble radius (MBR) from the recorded images was calculated. A red LED with a wavelength of 633 nm was used as a light source to illuminate the prism. All measurements were performed under red light only to avoid light induced changes of the sample. Foam properties were determined 3-fold.

Sum-Frequency Generation (SFG) Spectroscopy

SFG measurements were performed with a home-built broadband SFG spectrometer that is described elsewhere.[37] A femtosecond IR beam (fwhm bandwidth >200 cm–1) and an etalon filtered vis beam at 800 nm wavelength (fwhm bandwidth <6 cm–1) were overlapped at the air–water interface at incidence angles of 60° and 50°, respectively. Spectra were measured in the frequency range of 2800 to 3800 cm–1 in seven steps and in the frequency range of 950–1250 cm–1 in two steps. The acquisition time in the range of C–H and O–H stretching vibrations (2800–3800 cm–1) varied between 3 and 5 min and in the range of S–O stretching vibrations (950–1250 cm–1) between 10 and 30 s depending on the SF signal strength. All sample spectra were referenced to the nonresonant SFG signal of a plasma cleaned polycrystalline gold surface in order to take intensity changes with IR frequency into account. Air–water interfaces were investigated using ssp (SF/vis/IR) polarizations, while ppp was used for the reference spectrum of Au. Kinetic SFG measurements in the range of 2800–3800 cm–1 were performed by scanning the broadband IR beam in three steps only. For that, the center wavelength of the IR beam was set to 2900, 3200, and 3500 nm.

UV/Vis Spectroscopy

Ultraviolet/visible (UV/vis) absorption spectra were recorded for wavelengths of 200 to 800 nm using a Cary 100 Scan UV/vis spectrophotometer (Agilent, US). One ml of sample solution was filled into a semimicro UV cuvette and spectra were taken directly after sample preparation as well as after 1, 2, and 7 days of aging. The latter helped to access the colloidal stability of the samples. More information is given in the Supporting Information. The method of continuous variations, also known as Job’s plot (details in the Supporting Information), was applied in order to resolve the stoichiometry of ion pair formation between SSY and C16TAB. Job’s plots have been used in previous work for the investigation of dye-surfactant binding interactions.[6,38−41] For a description of the working principle and experimental results from the application of the Jobs method the reader is referred to the Supporting Information.

Experimental Results

Results from UV/Vis Spectroscopy

UV/vis spectra of bulk solutions were recorded for different Sunset Yellow/C16TAB molar ratios, where the SSY concentration was kept constant while the C16TAB concentration was varied. Three concentration series were investigated with 8, 15, and 30 μM Sunset Yellow. Figure a shows absorbance spectra of solutions with 15 μM SSY but different surfactant concentrations. The absorption bands at 480 and 314 nm decrease with increasing C16TAB concentration, while a shoulder is emerging at 425 nm indicating an increasing absorption band centered near this wavelength. As we will discuss below, these changes in UV/vis spectra are caused by electrostatic and hydrophobic interactions between anionic SSY2– and cationic C16TA+ ions.

Figure 2

(a) UV/vis absorbance spectra for different C16TAB bulk concentrations at a fixed SSY bulk concentration of 15 μM. (b) Decrease of absorbance at 480 nm in % as a function of the surfactant/azo dye–mixing ratio compared to the absorbance at 480 nm of pure SSY solutions for bulk SSY concentrations of 8, 15, and 30 μM. The solid black line in part b guides the eye.

In Figure b we compare the decrease of the absorption band at 480 nm in the presence of different C16TAB concentrations to the absorbance of Sunset Yellow solutions without surfactant additions. Increasing the C16TAB concentration causes a substantial decrease in absorbance, which saturates at a C16TAB/SSY molar ratio of ∼4. At this point, the absorbance at 480 nm has decreased by ∼45% as compared to the blank SSY solution.

The observed strong changes in absorbance spectra are indicative to a complete modification of Sunset Yellow molecules in terms of electronic and likely also molecular structure changes that originate from the electrostatic and hydrophobic interactions of SSY molecules with C16TAB surfactants. This conclusion is further corroborated by our surface tension and SFG measurements that are reported and discussed below. In Figure b, we also show that this behavior is independent of the bulk SSY concentrations between 8 and 30 μM because identical changes are observed in analogous experiments with different SSY concentrations. In order to resolve the stoichiometry of the binding between dye and surfactant, Job plots were performed, where the differential absorbance ΔA as a function of SSY volume fraction was recorded. The results, which are shown in the Supporting Information, support our conclusion that predominantly 1:2 ion pairs of SSY and C16TAB are forming.

Results from Tensiometry

Figure a presents the dynamic surface tension σ of SSY/C16TAB mixtures with the C16TAB concentration fixed to 120 μM. For the blank C16TAB solution an equilibrium surface tension of 67 mN/m is observed, while all other samples which contained SSY and C16TAB reach surface tensions around 36 mN/m but at different times after the air–water interface from these solutions was created. The time until an equilibrium state is reached decreased nearly exponentially with the SSY concentration as shown in Figure b. At dye concentrations above 60 μM, the samples were already in equilibrium within a few seconds and the initial stages in surface adsorption could not be resolved due to the limited time resolution of the pendent drop method at very short adsorption times. The latter is limited by the time needed to create the pending drop. The surface tension of samples with 4–15 μM SSY are very similar shortly after the surface was created and are close to the surface tension of blank 120 μM C16TAB solution. Obviously, the surface excess of all possible surface active species at short adsorption times is relatively low and at these times the surface is likely to be covered by free C16TAB molecules only. In fact, this conclusion can be corroborated by our time-resolved SFG spectra as we will show below. At later stages of the adsorption process other species such as SSY/C16TA+ ion pairs or the interaction of SSY with C16TAB at the air–water interface reduces the surface tension dramatically from ∼67 to 36 mN/m (detailed discussion below).

Figure 3

(a) Surface tension σ as a function of surface age t for different bulk SSY concentrations. (b) Time until equilibrium t is reached as a function of bulk SSY concentration. The solid line in part b guides the eye. The bulk C16TAB concentration was fixed to 120 μM.

Results from Foam Characterization

At a C16TAB concentration fixed to 120 μM, foam capacities decrease with increasing concentrations of Sunset Yellow in the bulk solution (Figure a). Although the foam stability is slightly improving when only small quantities of dye are added (<1 μM), higher concentrations of SSY lead to a marked decrease in the foam stability which is reduced to negligible values when the SSY concentration is >60 μM (Figure a).

Figure 4

(a) Foam stability after 30 min and foam capacity and (b) mean bubble radius (MBR) directly after foaming as a function of bulk SSY concentration at a fixed C16TAB concentration of 120 μM. Lines are a guide to the eye. At a bulk concentration of ≥60 μM no stable foams could be produced.

In addition, the foam structure (Supporting Information) also shows clear changes with rising SSY concentrations. For blank C16TAB solutions, wet foams are formed with quite narrow bubble size distributions of mainly small bubbles. With increasing SSY concentration the bubbles get larger and the size distribution broadens which is accompanied by a loss in foam stability as shown in Figure . Because of the poor foamability for bulk concentrations above 60 μM, no foam structure was detected as the foam did not reach the imaged section of the prism (Experimental Results).

Results from SFG Spectroscopy

Air–Water Interfaces in Equilibrium

SFG spectra were recorded at the air–water interface for SSY/C16TAB mixtures in equilibrium. Here, the surfactant concentration was kept constant at 120 μM and the dye concentration was varied between 0 and 120 μM. SFG spectra for different dye/surfactant mixing ratios are presented in Figure for (a) the frequency region where C–H and O–H stretching vibrations can be observed (2800–3800 cm–1) and (b) in the fingerprint region between 950 and 1350 cm–1.

Figure 5

Vibrational SFG spectra (a) in the range of 2800–3800 cm–1 (C–H and O–H stretching region) and (b) in the range of 950–1250 cm–1 (S–O stretching region) for different bulk SSY concentrations. The bulk C16TAB concentration was fixed to 120 μM. Red lines represent the fits to the experimental data as explained in the main text.

In the fingerprint region an intense vibrational band around 1115 cm–1 together with an overlapping weaker and much broader band that is centered around 1150 cm–1 are dominating the SFG spectra. Additional vibrational bands are observed at 980, 1026, and 1210 cm–1.

Vibrational bands centered at 980, 1026, and 1115 cm–1 can be assigned to vibrations of the azobenzene backbone of SSY,[42−44] whereas the broad band at 1150 cm–1 is attributable to the out-of-phase stretching vibration of C(aromatic ring)–N(azo group) bonds.4[45] Furthermore, we attribute the band at 1210 cm–1 to S–O stretching vibrations.[45] As a consequence, all bands observed in the fingerprint region are due to the presence of SSY moieties at the interface. At 2850 and 2880 cm–1 vibrational bands from CH2 and CH3 symmetric stretching bands as well as a CH3 Fermi resonance at 2929 cm–1 are clearly observed. Note that the CH3 Fermi resonance is due to a coupling of the CH3 symmetric stretching mode with the overtone of the CH3 bending mode at 1445 cm–1.[46] As only C16TA+ moieties at the interface contain methyl groups, the symmetric CH3 stretching band at 2880 cm–1 is indicative to the presence of C16TA+ surfactants at the interface, while the aromatic C–H stretching band at ∼3060 cm–1 originates from interfacial SSY molecules only.[47] Broad bands centered around 3200 and 3450 cm–1 arise from O–H stretching vibrations of interfacial H2O molecules.[28,47,48] Additionally, increasing the SSY concentration causes a decrease of O–H vibrational bands and leads to a strong nonresonant contribution χ(2) to the second-order electric susceptibility, which gets more intense as the surface excess of SSY moieties at the interface increases.

From a close inspection of Figure b, it becomes obvious that all vibrational bands increase in intensity with increasing SSY concentration. This is consistent with the increase in χ(2) and indicative to a higher surface excess of SSY at the air–water interface. In Figure , we compare the amplitude A of the 1115 cm–1 band with the changes in χ(2). The latter two were determined by nonlinear least-squares fitting of the SFG spectra in Figure a and 5b according to eq . Obviously, there is a close match between the changes in χ(2) and the change in SF amplitude of the 1115 cm–1 band with SSY concentration. For c(SSY) < 30 μM, a sharp increase in SF amplitude and χ(2) is observed, whereas in the c(SSY) region >30 μM both A and χ(2) reach plateau values. From a close analysis of these results, we conclude that at ∼60 μM a maximum in SSY surface excess is reached.

Figure 6

(a) SFG amplitude of the vibrational bands of water molecules. (b) Black (■): SFG amplitude of vibrational band of azo benzene backbone vibrations of Sunset Yellow (1115 cm–1) as a function of bulk SSY concentration. Red (●): Nonresonant part of the second order electric susceptibility. Lines are a guide to the eye. The bulk C16TAB concentration was fixed to 120 μM.

Figures a and 6a clearly show that the intensity of the water bands is decreasing with SSY concentration. At this point, we recall the strong relationship between SF signals of water bands and surface charging as explained in the Principles of Sum-Frequency Generation, eqs and 4. Consequently, we associate the observed behavior to a reduction in the surface net charge with increasing SSY bulk concentration. For pure C16TAB solutions, the air–water interface is dominated by positively charged C16TA+ surfactants only. This is changed after addition of SSY, where also SSY moieties are observed at the interface (Figures and 6). Since SSY molecules are not surface active at all, SSY moieties at the air–water interface must be closely related to the presence of C16TA+ ions. Presumably, hydrophobic and electrostatic interactions are the driving forces for SSY/C16TA+ ion pair formation that are highly hydrophobic and surface active.

Air–Water Interfaces under Nonequilibrium Conditions

The adsorption process of Sunset Yellow and C16TAB at the air–water interface was additionally investigated with SFG using a SSY/C16TAB mixture with 15 and 120 μM concentrations, respectively. Figure a presents a kinetic series of SFG spectra that were recorded shortly after the air–water interface was created (1 min). Obviously, there are substantial changes in the overall intensity and shape of the SFG spectra over time. At 1 min, the SFG spectrum is similar to the SFG spectrum of a pure C16TAB solution (Figures a and 7a), but with an already reduced intensity of the O–H contributions.

Figure 7

(a) Kinetic series of vibrational SFG spectra in the C–H and O–H stretching region at the air–water interface as a function of surface age. The first spectrum was recorded 1 min after the surface was created. (b) SFG amplitude of the aromatic 3060 cm–1 vibrational band and the nonresonant part of the second-order electric susceptibility as a function of surface age. Lines are a guide to the eye. The aqueous solutions had concentrations of 15 μM SSY and 120 μM C16TAB.

For a better understanding of the complex changes in SFG intensity and spectral shape, we provide in Figure b a quantitative analysis based on nonlinear least-squares fits to the experimental data in Figure a. Between 1 and 10 min, both the χ(2) contribution as well as the aromatic C–H stretching band at 3060 cm–1 arise and get more pronounced with surface age.

In fact, Figure a shows that for the concentrations (15 μM SSY and 120 μM C16TAB) where the kinetically resolved SFG measurements were performed, the equilibrium surface tension of 37 mN/m is also reached after ∼40 min. However, an analysis of our SFG spectra can add substantial new information to the results from tensiometry because the molecular identity of the absorbing species is directly resolved at specific adsorption times. Further analysis of the SFG spectra shows that the increase in both χ(2) and amplitude of the 3060 cm–1 band is accompanied by a decrease in SFG intensity of the O–H stretching bands. This points to a loss of polar order in the interfacial layer of water molecules and thus to a decrease of surface net charge (see Discussion below).

Discussion

Ion Pairing and Surface Charging

From the results of UV/vis measurements (Figure ), tensiometry (Figure ) and SFG spectroscopy (Figures and 7), we can now derive information on the interaction between Sunset Yellow and C16TAB as well as on the composition and structure of the air–water interface in and outside equilibrium conditions.

In addition, we can derive structure–property relations that link the interfacial molecular structures and interactions to macroscopic foam stability and structure. This is possible because aqueous foam is an inherently interface-controlled material.

By adding C16TAB, the extinction spectrum of Sunset Yellow molecules in water changes remarkably. The absorption bands at 480 and 314 nm decrease, while a shoulder is emerging at 425 nm and points to an absorption band centered near this wavelength.

The influence of cationic surfactants on the absorption spectrum of azo dyes was already studied in earlier work.[5−12] Although it is commonly accepted that the changes must be due to electronic and or structural changes to the best of our knowledge there is no clear consensus about the exact nature of the interactions and the molecular structures that are formed. In particular, changes in isomerization of the trans-azo dye to the cis-azo dye, as well as an increase of the hydrophobic interactions between the alkyl chain of the surfactant and the azo dye were attributed to the spectral changes.[9,10] Other studies proposed the formation of ion pairs consisting of anionic azo dye and cationic surfactants and also the formation of premicelles from these ion pairs below the surfactant critical micelle concentration (CMC).[11,12,49] In the latter case, the authors concluded that trans-azo dye molecules undergo cis-isomerization during micelle formation which is thought to minimize steric strain, as the cis-isomer is smaller in volume compared to the trans-isomer.[12]

However, independent of the exact nature of the formed structures after the ionic azo dye interacts with the ionic surfactant, it was concluded that the dramatic changes in UV/vis spectra and in particular of the absorption band at a wavelength of 480 nm can only be only due to complex or ion pair formation.[38,39,41] That is because such substantial changes demand strong electrostatic interactions. For that reason, we attribute the decrease of the absorbance band to the formation of SSY/C16TAB ion pairs and the plateau in Figure b at SSY: C16TAB molar ratios >1:2 to a 1:2 stoichiometry of anionic SSY and cationic C16TA+ ions ([SSY]2–[C16TA+]2). A Job’s plot that is presented among with further details on this method in the Supporting Information brings strong support to this conclusion.

In addition, our conclusion of ion pairing with 1:2 stoichiometry is consistent to the substantial decrease in surface tension to equilibrium values of 35 mN/m (Figure ), because nonionic surfactants such as [SSY]2–[C16TA+]2 ion pairs have a higher efficiency to reduce the surface tension than ionic surfactants.[50] Further evidence comes from our analysis of O–H vibrational bands around 3200 and 3450 cm–1 (Figures and 6): For molar ratios >1:2 the highest surface coverage of SSY moieties is reached (Figure ), but is accompanied by the absence of strong O–H bands. This clearly shows that the air–water interface at these conditions is uncharged and brings further evidence to our conclusion on [SSY]2–[C16TA+]2 pairs.

However, a maximum in surface excess is already achieved at SSY concentrations <30 μM which is below the concentration where the surface charge is fully compensated. This implies that at intermediate concentrations between 30 and 60 μM where still free C16TAB molecules are available in the bulk, the latter can coadsorb to [SSY]2–[C16TA+]2 ion pairs and cause a positive net charge at the air–water interface. In addition to the equilibrium properties of SSY and C16TAB modified air–water interfaces, also information on nonequilibrium properties from SFG and tensiometry need to be considered. Looking at the results from tensiometry (Figure ) and SFG (Figure ) it becomes clear that at early stages of the adsorption process free C16TA+ molecules dominate the air–water interface.

This can be derived from a close inspection of the kinetic series of SFG spectra in Figure which demonstrates that the SFG spectrum at early stages of the adsorption process is very similar to the SFG spectrum of a blank C16TAB solution. In addition, there is no signature of SSY moieties at the interface as both amplitudes of the nonresonant and the aromatic C–H stretching contributions are negligible at very early adsorption times. This is consistent with the value of the initial surface tension (for adsorption times <1 min) e.g. 67 mN/m for solutions of 8 μM SSY and 120 μM C16TAB being close to the initial value of 68 mN/m for blank 120 μM C16TAB solution. From these observations we can now conclude that [SSY]2–[C16TA+]2 ion pairs adsorb much slower than free C16TA+ ions. For longer adsorption times, the aromatic C–H-stretching and the nonresonant contribution arise in our SFG spectra and get more pronounced over time. This is accompanied by a decrease in the SFG intensity of O–H stretching bands.

Surface Charging and Foam Properties

For SSY concentrations <30 μM, foam stability and foam structure are hardly affected by the rise in Sunset Yellow concentrations (Figure ). This is changed for concentrations >30 μM where the foam stability and the foam capacity are both decreasing while the mean bubble size is considerably increasing. Results from tensiometry (Figure ) and SFG spectroscopy (Figure ) can help to explain these observations.

At SSY concentrations <30 μM, [SSY]2–[C16TA+]2 ion pairs adsorb very slowly (see Discussion, above). Since it takes up to several hours to reach the equilibrium state, the foams are mainly stabilized by free C16TA+ ions.

Above 30 μM SSY, [SSY]2–[C16TA+]2 ion pairs are at the interface already within a few seconds and can contribute to the foam properties significantly. Although the ion pairs lower the surface tension very efficiently, the decrease in surface tension does not lead to an increase but to a decrease in foam stability.

In order to explain this behavior, we need to consider also the results from our SFG spectra which provide additional information on interfacial charging that is obtained by the changes in O–H amplitude with SSY concentration. At this point, we recall that electric field-induced contributions to the SF intensity can substantially change the intensity of O–H vibrational bands as a function of the interfacial electric field (see Principles of Sum-Frequency Generation).[33,34] In fact, higher SSY concentrations cause not only lower surface tensions but also result in a loss of net charge at the air–water interface and thus in a decrease in electrostatic repulsion within the foam films. As a consequence of the decrease in interfacial net charge, the electrostatic stabilization of foam films (lamellae) and the macroscopic foam weakens and leads to a loss in foam stability. This is in particular true for SSY concentrations >60 μM, where the SFG intensity from interfacial water is close to zero values and points to an uncharged interface. These observations are in excellent agreement with the classical theory of electrostatic stabilization of foam films and the electrostatic disjoining pressure.[51−54]

Conclusions

In this work, information from vibrational SFG spectroscopy, tensiometry and UV/vis spectroscopy was successfully combined to identify molecular building blocks at air–water interfaces and to understand the adsorption process of azo dye/surfactant ion pairs at the air–water interface. From an analysis of macroscopic foam, we were able to derive structure–property relations between interfacial building blocks, surface charging and foam properties.

Interfaces dominated by C16TA+ ions result into relatively stable foams, while an increase in SSY concentrations causes both foamability and stability to decrease. This originates from adsorption of [SSY]2–[C16TA+]2 ion pairs at the interface, which decrease the surface tension considerably but also lead to a decrease in surface net charge. Consequently, the decrease in stability of macroscopic foam with SSY concentration at a fixed C16TAB concentration can be explained by a loss of electrostatic repulsion forces between the two opposing interfaces in foam films. Since these are a major hierarchical element in macroscopic foam, their properties on a molecular scale have dramatic effects on the macroscopic scale.