924
Langmuir 2006, 22, 924-932
Effects of Nonionic Surfactant C12E5 on the Cooperative Dynamics of Water Simon Schro¨dle,† Glenn Hefter,‡ Werner Kunz,† and Richard Buchner*,† Institut fu¨r Physikalische und Theoretische Chemie, UniVersita¨t Regensburg, D-93040 Regensburg, Germany, and Chemistry - DSE, Murdoch UniVersity, Murdoch, W.A. 6150, Australia ReceiVed July 21, 2005. In Final Form: NoVember 11, 2005 A dielectric relaxation study of binary mixtures of nonionic surfactant C12E5 + water has been made as a function of temperature in the isotropic micellar, lamellar, and hexagonal regions of the phase diagram. Two dielectric dispersion steps were found and could be assigned to the intermolecular cooperative dynamics of water at the micellar interface and in the bulk water domains. A quantitative analysis is given. The relaxation amplitudes were used to determine effective hydration numbers. The activation energies of water relaxation were calculated from the relaxation times. The data indicate weaker surfactant-water and water-water interactions near the micellar interface compared to those of bulk liquid water. Further analysis revealed the presence of water clusters large enough to show a cooperative relaxation mode even at high surfactant concentrations. However, the relaxation time of this mode is larger compared to that of pure water. This points out the importance of confinement effects on water dynamics.
1. Introduction Nonionic surfactant solutions and microemulsions play an increasingly important role in industry and are also a focus of basic research. Typical representatives for this class of substances are the oligo(ethylene glycol) monoalkyl ethers, H2n+1Cn(OCH2CH2)mOH, commonly abbreviated as CnEm. Alterations of the lengths of the alkyl and/or glycol chains produce wide variations in the phase behavior of the aqueous solutions of these compounds. A number of isotropic and liquid-crystalline phases are known to exist for such mixtures, and transitions occur readily by changing the composition, temperature, or pressure of the system.1-5 From the strong temperature dependence of the observed physicochemical properties, it is obvious that nonionic surfactant solutions and their microemulsions reflect a delicate balance of thermal motions and the interaction forces between surfactant and water molecules. Despite a significant number of investigations including techniques such as calorimetry,7 Raman,8 UV,9 infrared,10 and NMR spectroscopy11-13 as well as scattering experiments,14,15 the molecular origin of the structural changes * To whom correspondence should
[email protected]. † Universita ¨ t Regensburg. ‡ Murdoch University.
be
addressed.
E-mail:
(1) Strey, R. Ber. Bunsen-Ges. Phys. Chem. 1996, 100, 182. (2) Schubert, K.-V.; Kaler, E. W. Ber. Bunsen-Ges. Phys. Chem. 1996, 100, 190. (3) Schick, M. J. Nonionic Surfactants: Physical Chemistry; Marcel Dekker: New York, 1987. (4) Mitchell, D. J.; Tiddy, G. J. T.; Waring, L.; Bostock, T.; McDonald, M. P. J. Chem. Soc., Faraday Trans. 1 1983, 79, 975. (5) Strey, R.; Schoma¨cker, R.; Roux, D.; Nallet, F.; Olsson, U. J. Chem. Soc., Faraday Trans. 1990, 86, 2253. (6) Schro¨dle, S.; Buchner, R.; Kunz, W. Fluid Phase Equilib. 2004, 216, 175. (7) Olofsson, G. J. Phys. Chem. 1985, 89, 1473. (8) Marinov, V. S.; Nickolov, Z. S.; Matsuura, H. J. Phys. Chem. B 2001, 105, 9953. (9) Christenson, H.; Friberg, S. E. J. Colloid Interface Sci. 1980, 75, 276. (10) Zheng, L.; Suzuki, M.; Inoue, T.; Lindman, B. Langmuir 2002, 18, 9204. (11) Rendall, K.; Tiddy, G. J. T. J. Chem. Soc., Faraday Trans. 1 1984, 80, 3339. (12) Nilsson, P.-G.; Lindman, B. J. Phys. Chem. 1984, 88, 5391. (13) Tonegawa, A.; Ohno, K.; Matsuura, H.; Yamada, K.; Okuda, T. J. Phys. Chem. B 2002, 106, 13211.
of these liquids is still not well understood. In particular, there is very limited knowledge of the cooperative dynamic processes occurring in these systems. Dielectric relaxation spectroscopy (DRS) is a powerful tool for the investigation of liquid-state dynamics and is especially sensitive to the cooperative motions in hydrogen-bonded systems such as water.16,17 Recent studies on anionic and cationic surfactant systems showed that broadband DRS provides otherwise inaccessible information on the structure and dynamics of micellar systems, especially on micelle hydration, which can be correlated with results from other methods.18-21 In continuation of our previous investigations, the purpose of the present article is to study the effect of a hydrophilic surface with uncharged H-bond acceptor sites on the water dynamics. The results for this model system are not only of interest for the understanding and preparation of artificial nanostructured systems but also have implications for protein chemistry and many biological systems in general. Figure 1 shows the phase diagram5 of the C12E5 + water binary system. The notation of Tiddy22 is used for the observed phases: L1, L2, and L3 are micellar, reverse micellar, and dilute micellar isotropic liquid phases, respectively; LR is a lamellar phase; and V and H are cubic and hexagonal liquid-crystalline phases, respectively. Also indicated in Figure 1 are the investigated state points, consisting of five series of measurements at constant (14) Lesemann, M.; Martı´n, A.; Belkoura, L.; Woermann, D.; Hoinkis, E. Ber. Bunsen-Ges. Phys. Chem. 1997, 101, 228. (15) Brunner-Popela, J.; Mittelbach, R.; Strey, R.; Schubert, K.-V.; Kaler, E. W.; Glatter, O. J. Chem. Phys. 1999, 110, 10623. (16) Feldman, Yu.; Skodvin, T.; Sjo¨blom, J. Dielectric Spectroscopy on Emulsion and Related Colloidal Systems - A Review. In Encyclopedic Handbook of Emulsion Technology; Sjo¨blom, J., Ed.; Marcel Dekker: New York, 2001; Chapter 6. (17) Buchner, R. Dielectric Spectroscopy of Solutions. In NoVel Approaches to the Structure and Dynamics of Liquids: Experiments, Theories and Simulations; Samios, J., Durov, V. A., Eds.; NATO Science Series II; Kluwer: Dordrecht, The Netherlands, 2004; Vol. 133. (18) Baar, C.; Buchner, R.; Kunz, W. J. Phys. Chem. B 2001, 105, 2906. (19) Baar, C.; Buchner, R.; Kunz, W. J. Phys. Chem. B 2001, 105, 2914. (20) Fernandez, P.; Schro¨dle, S.; Buchner, R.; Kunz, W. ChemPhysChem 2003, 4, 1065. (21) Buchner, R.; Baar, C.; Fernandez, P.; Schro¨dle, S.; Kunz, W. J. Mol. Liq. 2005, 118, 179. (22) Tiddy, G. J. T. Phys. Rep. 1980, 57, 1.
10.1021/la0519711 CCC: $33.50 © 2006 American Chemical Society Published on Web 12/22/2005
Effects of C12E5 on CooperatiVe Dynamics of Water
Langmuir, Vol. 22, No. 3, 2006 925 F)
[ (
1 1 1 +w 0- 0 F0w F2 Fw
)]
-1
(1)
Because the experimental F02 values are well fit by the equation F02(T) ) [991.66 - 0.99(T/K - 273.15)] kg m-3
Figure 1. Phase diagram of the system water/C12E5.5 Closed symbols indicate broadband dielectric relaxation measurements using TDR, VNA, and IF instruments. Only VNA data were recorded for points with open symbols (see text).
composition and variable temperature. Three paths were confined to the isotropic micellar phase corresponding to dilute surfactant concentrations (weight fractions of C12E5: w ) 0.05, 0.10, 0.20). With the sample of w ) 0.40, the H/L1 and L1/LR phase boundaries were crossed. Measurements for the most concentrated solution (w ) 0.70) were restricted to the lamellar phase. Additionally, a series with w ) 0.10 in D2O was performed to check possible isotope effects on the solvent dynamics. 2. Experimental Details and Data Analysis 2.1. Materials. All experiments were performed using C12E5 (Nikkol, Japan) of >99.8% purity (GC). Heavy water mixtures employed D2O with g99.9% isotope enrichment (Euriso-top, France) and were used without further purification. Aqueous samples were prepared with Millipore (Milli-Q) water from degassed components several days before use and then kept at the measurement temperature to ensure equilibration. For all experimental paths, the positions of the relevant phase boundaries (Figure 1) were checked with the automated instrument described previously.6 Note that the surfactant, both in its pure liquid state and in aqueous solution, shows significant degradation when exposed to air. Decomposition noticeably manifests itself in a decrease of the cloud point by several Kelvin. Therefore, samples were stored in the dark under N2(g).23 Benzonitrile, propylene carbonate, N,N-dimethylacetamide (DMA), and N,N-dimethylformamide (DMF), required for the calibration of the dielectric spectrometers, were of analytical grade (Merck, Germany). For the calculation of molar concentrations, c/mol L-1, essential for the analysis of the dielectric spectra, the densities, F, of the binary mixtures were required. These were measured using a vibrating-tube densimeter (Paar DMA60/601HT) calibrated with N2(g) and water. However, because of the high viscosity of the surfactant-rich solutions and the cost of C12E5, this was not possible for all mixtures at all temperatures. Estimates of the missing data were obtained from the known densities of water, F0w,24 and of the pure surfactant, F02,25 assuming ideal mixing behavior: (23) The rather expensive surfactant can be recovered from used solutions by liquid/liquid extraction with methylene chloride, careful drying with anhydrous sodium sulfate, and subsequent purification using a high-vacuum molecular distillation apparatus at ∼10-6 mbar. (24) The International Association for the Properties of Water and Steam. Release on the IAPWS Industrial Formulation 1997 for the Thermodynamic Properties of Water and Steam; IAPWS Meeting, Erlangen, Germany, Sept 1997.
(2)
in the temperature range between 25 and 50 °C, eq 2 was used to estimate densities for the supercooled surfactant at temperatures 0 indicates for the water molecules interacting with the oligo(ethylene glycol) segments a symmetric distribution of states (environments) that interconvert on a time scale Jτ1, in contrast to samples with w e 0.2 where thermal equilibration is rapid. Figure 10 shows that compared to solutions with w e 0.2 the relaxation time τ2 of bulklike water is now significantly longer than the corresponding value of pure H2O. Nevertheless, the temperature dependence of this mode is very similar (Figure 11). The activation enthalpy, ∆Hq298, is slightly increased, but ∆Sq298 and ∆Cqp do not differ from those of pure water (Table 4). Also, the (S2, τ2) mode neither shows a distribution of relaxation times nor reflects that two phase boundaries are crossed for the sample with w ) 0.4, whereas at w ) 0.7 the solution is entirely in the lamellar phase (Figure 1). Note that τ2 is not correlated with the (60) Garde, S.; Yang, L.; Dordick, J. S.; Paulaitis, M. E. Mol. Phys. 2002, 100, 2299. (61) Sterpone, F.; Pierleoni, C.; Briganti, G.; Marchi, M. Langmuir 2004, 20, 4311.
Effects of C12E5 on CooperatiVe Dynamics of Water
Figure 10. Ratio τ2/τw of the dielectric relaxation times of bulklike, τ2, and pure water, τw, as a function of surfactant weight fraction, w, at 288.15 K (b) and 298.15 K (9).
Langmuir, Vol. 22, No. 3, 2006 931
Figure 12. Temperature dependence of the relative amplitudes, Si/(S1 + S2), of bound (S1, full symbols) and bulklike (S2, open symbols) water of aqueous C12E5 solutions: w ) 0.40 (b, dashed lines); w ) 0.70 (9, dotted lines).
w ) 0.7, where all data are confined to the LR phase. Because of the scatter of the data, only a two-parameter fit is possible. Compared to the isotropic micellar solutions, ∆Hq298 and ∆Sq298 are significantly increased, and the activation entropy is now positive, suggesting either more and/or stronger hydrogen bonds to break before relaxation can take place.
Figure 11. Arrhenius plot of the relaxation times of bound (τ1, full symbols) and bulklike (τ2, open symbols) water in aqueous C12E5 solutions with w ) 0.40 (b) and w ) 0.70 (9). Also shown is the relaxation time of the cooperative relaxation mode of pure water (broken line).
viscosity of the samples, which increases by orders of magnitude when passing from the isotropic micellar phase (L1) to the liquidcrystalline H and LR phases. These observations clearly support the interpretation of τ2 as a dynamical property that reflects interactions occurring at distances that are short compared to the size of the supramolecular aggregates (H: three-dimensional array of cylindrical micelles; LR: stacked lamellae) but large enough to be similar to the relaxing domain in pure water. Because it is unlikely that the interactions of bulklike and bound water change significantly with w, the significant increase of τ2 with w (Figure 10), may reflect confinement effects on the water dynamics similar to the observations of Kremer et al. for water confined in pores.63 Compared to solutions of low surfactant content, the relaxation time of bound water changes significantly at w ) 0.4 and 0.7. The temperature dependence is even more pronounced, with a clearly upward curvature in the Arrhenius plot (Figure 11). A fit of log τ1 versus 1/T with eqs 5 and 6 is only reasonable for (62) Vedamuthu, M.; Singh, S.; Robinson, G. W. J. Phys. Chem. 1996, 100, 3825. (63) Kremer, F.; Huwe, A.; Arndt, M.; Behrens, P.; Schwieger, W. J. Phys.: Condens. Matter 1999, 11, A175.
For w ) 0.4, Figure 11 suggests that at low T the change in τ1 is similar to the w ) 0.7 solution despite them being two different phases. With increasing temperature, d log τ1/d(1/T) decreases markedly so that at 328 K the relaxation times for bound water in both w ) 0.4 and 0.7 solutions are identical, as might be anticipated because the phase is the same (LR, Figure 1). This shows the large influence of geometric constraints and (probably) headgroup mobility on the relaxation times τ1 of water incorporated into the interfacial regions and suggests that within the lamellar phase water-solute interactions are very similar. The relaxation time τ1 does not show marked changes at the phase boundaries of the w ) 0.4 sample. This means that on the length scale probed by the slow mode the environment does not change abruptly on phase transition but rather gradually. Interestingly, over the entire temperature range, τ1 was found to be shorter at w ) 0.40 than in the more dilute solutions. This may be connected with the decreasing curvature of the micellar interface,3 which finally leads to the formation of the lamellar structure. Figure 12 shows the temperature dependence of the relative amplitudes, Si/(S1 + S2), of bound and bulklike water for C12E5 solutions with w ) 0.4 and 0.7. For the former, both amplitudes are approximately equal, indicating that a significant amount of bulklike water is still present. However, for w ) 0.7 almost all water is bound, especially at low temperatures. For both solutions, the low-frequency contribution, S1/(S1 + S2), decreases linearly with increasing temperature, as expected by the progressing dehydration of the headgroups also observed with other techniques.51 Note that the slopes of S1/(S1 + S2) versus T are equal for both w ) 0.4 and 0.7. This suggests that the difference in the energies of bound and bulklike water is also the same. Although a quantitative analysis of the amplitudes is not feasible because of the likely change in the effective dipole moment for S2, it appears that the interaction of water with the oxyethylene groups is stronger compared to that of the dilute samples.
932 Langmuir, Vol. 22, No. 3, 2006
Conclusions The dielectric properties of C12E5 + H2O (or D2O) binary mixtures in the isotropic micellar, lamellar, and hexagonal phase regions were studied. Two dielectric dispersion steps were resolved and have been assigned to the intermolecular cooperative dynamics of bound water in the interface region (amplitude S1, relaxation time τ1) and to bulk water domains (S2, τ2). The activation enthalpy of the τ2 process is almost independent of the C12E5 concentration, w, and is similar to the value of pure water, ∆Hq298(H2O) (Table 4). This indicates that even at the highest surfactant concentration the relaxation cluster producing this mode is sufficiently large to behave as bulk water. However, the significant increase observed for τ2 at w g 0.4 (Figure 10) suggests that confinement effects are becoming important. The activation enthalpy of the τ1 process depends on w and is smaller than ∆Hq298(H2O) at w e 0.2 but larger for w ) 0.7. This suggests that at low surfactant concentrations the surfactantwater interactions and possibly also the water-water interactions in the hydration shell are weaker than the water-water interactions within the bulklike H2O. The finding τ1 > τ2 is a consequence of the unequal ratio of H-bond donor and acceptor positions and
Schro¨dle et al.
of the cooperative motions of solvent and surfactant headgroups required for relaxation of the hydration water. These effects are also relevant at w g 0.4, but now ∆Hq298(H2O) also suggests stronger intermolecular interactions within the hydration layer. From a quantitative analysis of the bulk water response, the effective hydration number of the oxyethylene groups was calculated for the water-rich mixtures (w e 0.2). The results obtained (Figure 7) are in good agreement with the scarce literature data and confirm progressive dehydration of the oxyethylene groups with rising temperature. Analysis of the slow relaxation process suggests that the effective dipole moment of the water molecules interacting with the surfactant headgroups is considerably lower than that of pure water, implying significant orientational correlations within the hydration shell that do not change with concentration and temperature. This suggests that dehydration and probably also phase separation are essentially entropy-driven. Acknowledgment. S.S. appreciates the support of the Verband der Chemischen Industrie e.V. (VCI). LA0519711
