In Fig. 1 we show HDVSFG spectra in the frequency range between 2750 and 3675 cm− 1 of an aqueous solution containing 700 nM SDS, an aqueous solution containing 70 µM C12E6, and an aqueous solution containing both 700 nM SDS and 70 µM C12E6. We present the imaginary component of the second-order susceptibility Im[χ(2)] measured for these solutions. The HDVSFG spectrum of water shows a weak broad negative band between 3200 and 3600 cm− 1, whereas the HDVSFG spectra of the solutions of SDS and C12E6 show two strong positive bands (3240 and 3400 cm− 1). These responses are all attributed to the O-H stretch vibrations of different hydrogen-bonded water molecules35,36. The negative sign of the broad OH signal observed for pure water indicates that the hydrogen-bonded OH groups of the water molecules have a net orientation towards the bulk. The positive sign of the OH signals observed for the surfactant solutions shows that the hydrogen-bonded OH groups have a net orientation away from the bulk, which for SDS can be well explained by the negative charge of the dodecyl sulfate (DS−) ions accumulated at the surface30,31. For C12E6, the net orientation of the water OH groups towards the surface results from the formation of hydrogen bonds to the ether oxygen of the headgroup of the surfactant.29
The HDVSFG spectra of the solutions containing C12E6 also show two strong negative features at 2850 and 2920 cm− 1 and a small positive band at 2960 cm− 1. The bands at 2850 and 2965 cm− 1 are assigned to the symmetric and antisymmetric stretching of the terminal CH3 group. The absence of any significant modes in between the two prior confirms the CH2 dipoles are fully canceled and there is no gauche effect in this case. The 2930 cm− 1 is assigned to the fermi-resonance between the symmetric CH3 stretching and overtones of the C-H bending modes of the CH3 group30,37. The positive sign of the asymmetric stretch band indicates that the aliphatic tails of the C12E6 surfactant molecules are pointing towards the air, away from the solution as expected.
A surprising and exciting feature in Fig. 1 is that the water signal of the solution containing both surfactants is much stronger than the added signal of the two solutions separately containing only C12E6 or SDS. The presence of C12E6 at CMC in the solution enhances the response of the water molecules to the addition of 700 nM SDS by a factor of ~ 10.
In Fig. 2 we show HDVSFG spectra of binary mixtures of SDS and C12E6 over a wide range of SDS concentrations. The HDVSFG spectra of SDS solutions only at the same concentrations as used in the binary mixture are reported in the supplementary information (Figure SI 1). We observe a steady increase in the intensity of the water signal with increasing SDS concentration in both ranges of SDS concentrations where C12E6 is in excess (Fig. 2a) or SDS is in excess (Fig. 2b). To check whether the observations may be influenced by heat-induced effects, we repeated the measurement at one of the intermediate concentrations with a much reduced infrared power. This did not change the results (Figure SI 2). The responses of the C-H vibrations of the packed monolayer of surfactants remain constant with SDS concentration. The apparent change in the amplitudes of the C-H vibrations with SDS concentration is due to the rise of the low-frequency wing of the response of the water O-H vibrations with increasing SDS concentration.
To study the synergetic effect of SDS and C12E6 quantitatively, we show in Fig. 3 the ratio of the enhanced water signal ranging between 3200 and 3600 cm− 1 induced by SDS both in the presence of 70 µm C12E6 and in the absence of C12E6, as a function of the concentration SDS. The synergistic effect was calculated using the following equation \(\frac{|{S}_{{H}_{2}O, bin. mixt. }- {S}_{{H}_{2}O, 70 \mu M {C}_{12 }{E}_{6}}| }{|{S}_{{H}_{2}O, SDS} - { S }_{{H}_{2}O, Pure {H}_{2}O}|}\) where “S” indicates the maximum amplitude of the water signal for different solutions. A ratio larger than 1 implies that there is a synergetic effect on the water signal. It is seen that the synergetic effect of SDS and C12E6 on the water signal is very strong for SDS concentration up to ~ 3.5 µM SDS (Fig. 3 inset), and vanishes (the ratio attaining a value of ~ 1) at an SDS concentration of ~ 70 µM. At SDS concentrations > 70 µM the ratio drops below 1, showing that at these concentrations the competition of the two surfactants for the limited surface area becomes more important than the synergetic effect.
We performed molecular dynamics simulations to identify the driving force behind the synergy of SDS and C12E6 on the response of the interfacial water molecules. We employed the umbrella sampling (US) technique (see Computational Methods) to compare the binding free energy due to DS− adsorbing to the bare air-water interface with that of DS− adsorbing to a surface fully covered with C12E6. Panels a, b, and c of Fig. 4 visually depict the three systems that are simulated, i.e. DS− adsorbing to the water-air interface, the water-C12E6-air interface (without DS−), and DS− adsorbing to the water-C12E6-air interface, respectively. From our simulations, we find that at low concentrations, interfacial DS− tends to orient nearly parallel to the surface when adsorbed to the water-air interface, and orient nearly perpendicular to the water surface when adsorbed to the water-C12E6-air interface, as depicted in panels a and c, respectively.
Figure 4d shows a comparison between the free energy profiles computed for each system as a function of the DS- distance from the interface. Zero distance corresponds to the Gibbs dividing interface where the water density has half its bulk value. We find that the binding free energy of DS- adsorbing to a bare water-air interface is -8 kBT, while that of DS- adsorbing to the water-C12E6-air interface is -19 kBT. This roughly 10 kBT enhancement of the binding free energy due to the presence of a C12E6 monolayer at the interface leads to a much higher surface concentration of DS- for a water surface that is covered with a layer of C12E6 than for the bare water surface.
The simulations allow for the determination of the entropic and enthalpic contributions of DS− binding to the water-C12E6-air interface, as shown in Fig. 4e (the decomposition of the entropic and enthalpic contributions of DS− binding to the air-water interface is shown in SI 3). The results show that the enhancement of the free energy for adsorption of DS− is largely enthalpic. The binding of DS− to the surface is favored enthalpically by roughly − 45 kBT while the entropic term incurs a penalty of approximately 25 kBT.
The observation that the binding of DS− to the water-C12E6-air interface is enthalpically driven, is somewhat surprising, as hydrophobic aggregation or self-assembly of aliphatic tails in aqueous environments is usually driven by a gain in entropy38,39. In fact, for DS− binding to a bare water surface (no C12E6), the adsorption is entropically driven (SI 3). The fact that the adsorption of DS− to a C12E6-covered water surface is enthalpy-driven can be explained as follows. Firstly, there is a drastic change in DS− orientation leading to favorable van der Waals interactions between the aliphatic hydrocarbon chains (Fig. 4c). Furthermore, previous work showed that C12E6 creates a 3 nm thick polarized layer of water at the interface29 which is not present in pure water. When DS− absorbs at the water-C12E6-air interface, this polarized water layer gets even thicker. This extended polarized water layer constitutes a significant attractive enthalpic contribution (SI 4). In addition, besides the role of the water, the simulations also suggest important contributions coming from changes in both the sodium counterion and C12E6 upon DS- binding to the surface. The extended orientation of the water molecules at the water-C12E6-air interface limits their conformational space, which largely explains the entropic penalty of approximately 25 kBT, associated with the adsorption of DS− to the interface.
In Fig. 4f the enhanced orientation of the water molecules induced by the adsorption of DS− to the water-C12E6-air interface is illustrated by calculating the integrated water dipole densities per unit volume as a function of the perpendicular direction to the interface (Z-coordinate). Note that the dipole is integrated from the air towards the bulk. The net dipole caused by the orientation of water clearly shows a very strong sensitivity to the presence of DS−. In agreement with the HDVSFG results illustrated in Figs. 1 and 2, Fig. 4f shows that adding SDS to the C12E6 at CMC leads to a significant enhancement of the orientation of the water molecules.
The present results demonstrate that the interaction between C12E6 and a negatively charged surfactant can be highly favorable, and can have a profound effect on the structure and orientation of nearby water molecules. In recent years the study of the interaction of C12E6 with other organic and inorganic molecules has emerged as an important research topic due to its wide range of applications40. For instance, it has been shown that nonionic surfactants form mixed micelles with ionic surfactants, and reduce the interaction between the ionic surfactant and the protein under consideration21. It has thus been found that the addition of a neutral surfactant to an SDS-protein complex reduces the protein denaturing capability of SDS, and promotes the refolding of different membrane proteins from the SDS-bound complexes41–43. The presently observed highly favorable interaction between C12E6 and SDS offers a potential explanation for this effect. Protein denaturation by SDS likely relies on the favorable interaction of the hydrophobic tail of the DS− ion and the hydrophobic residues of the protein. When adding a neutral surfactant like C12E6 to complexes of DS− and unfolded proteins, the favorable interaction between the hydrophobic tails of the DS− ions and the hydrophobic residues will likely be replaced by the even more favorable interaction between the hydrophobic tails of C12E6 and SDS. As a result, the hydrophobic protein residues may detach from DS− and re-aggregate, implying a (partial) refolding of the protein.