Langmuir 2006, 22, 11205-11207
11205
Pyrene Fluorescence at Air/Sodium Dodecyl Sulfate Solution Interface Robin Humphry-Baker,† Michael Gra¨tzel,† and Yoshikiyo Moroi*,‡ Institute de Chimie Physique II, Ecole Polytecnique Federal de Lausanne, CH-1015 Lausanne, Switzerland, and Graduate School of Pharmaceutical Sciences, Kyushu UniVersity, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan ReceiVed July 12, 2006. In Final Form: October 6, 2006 The dependence of pyrene fluorescence spectra on the concentration of sodium dodecyl sulfate (SDS) was observed, where the solution was prepared from water saturated with pyrene. The values of the I1/I3 ratio from the bulk solution and from the upper meniscus region in an optical cell were similar but decreased rapidly around the critical micelle concentration (cmc) of SDS, indicating that pyrene molecules preferred to be solubilized in the micelles having a lower dielectric constant. The fluorescence intensity of the excimer indicated the concentration of pyrene molecules at the air/solution interface or the surface activity of pyrene molecules. In addition, the intensity from the meniscus region is much larger than that from the bulk at the concentrations below the cmc, whereas there was no difference in the intensity between the bulk and the meniscus above 8 mmol dm-3 of SDS. The analysis of the fluorescence intensity from the excimer strongly suggests the presence of molecular aggregates that are favorable to the pyrene molecules just like the micelles in the bulk, making them less movable.
Introduction The surface tension of an aqueous surfactant solution is a thermodynamic variable that is specified by the structural interaction between the surfactant molecules and the water molecules as well as the multiple molecular layers beneath the air/solution interface. That is, the variation in surface tension of a soluble surfactant solution should depend on the change in the steric structure of the many molecular layers beneath the air/ solution interface. According to the conventional Gibbs adsorption model, the adsorbed film of soluble amphiphiles locates at the air/solution interface just like an insoluble monolayer, which is illustrated in many text books on colloid and interface science.1 However, in recent years, one of the authors has questioned the conventional concept of the Gibbs adsorbed film, on the basis of the water evaporation rate from aqueous soluble surfactant solutions.2 Accordingly, a new concept of the Gibbs surface excess has been presented. Fortunately, this new concept was also supported by Brewster angle microscopy (BAM) images,2 surface tension,3 evaporation from liquid mixture,4 and surface potential.5 The effect of insoluble monolayers on the rate of water evaporation was carefully investigated. The retardation of the evaporation was found to be due to the presence of a monolayer just at the air/water interface.6 The insoluble molecules do not appear to disturb the steric structure formed by water molecules below the interface, whereas soluble surfactants should destroy the steric structure because of their rapid motion up and down * Corresponding author. Address: Graduate School of Pharmaceutical Sciences, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan. Fax: +81-92-606-1536. E-mail: moroiscc@ mbox.nc.kyushu-u.ac.jp. † Ecole Polytecnique Federal de Lausanne. ‡ Kyushu University. (1) Adamson, A. W.; Gast, A. P. Physical Chemistry of Surfaces, 6th ed.; Interscience: New York, 1997. Hiemenz, P. C. Principles of Colloids and Surface Chemistry; Marcel Dekker: New York, 1986. (2) Moroi, Y.; Rusdi, M.; Kubo, I. J. Phys. Chem. B 2004, 108, 6351 (3) Rusdi, M.; Moroi, Y.; Hlaing, T.; Matsuoka, K. Bull. Chem. Soc. Jpn. 2005, 78, 604. (4) Rusdi, M.; Moroi, Y.; Nakahara, H.; Shibata, O. Langmuir 2005, 21, 7308. (5) Nakahara, H.; Shibata, O.; Moroi, M. Langmuir 2005, 21, 9020. (6) Rusdi, M.; Moroi, Y. J. Colloid Interface Sci. 2004, 272, 472.
around the interface with a lifetime of about a microsecond. This assumes that the surfactant molecules are truly located just at the air/solution interface. Nevertheless, the soluble surfactants have been shown not to change both the water evaporation rate and its activation energy at their different concentrations.2 In other words, there is no difference in the water evaporation rate and its activation energy between a surfactant solution and purified water. In addition, the images of BAM for soluble surfactant solutions are very similar to that of pure water itself.2 If any miscible component other than the water molecules truly exits just at the air/solution interface, the evaporation rate would become nearly proportional to the mole fraction at the surface, because the evaporation must take place just from the interface.4 According to the new idea, an adsorbed film is composed of double molecular layers, just like a cell membrane, and locates at a certain distance below the air/solution interface. The double layer was deduced from the change in surface tension with concentration or the Gibbs surface excess,3 in which the surface tension versus concentration curve for soluble surfactant solution was divided into three regions in order to solve the Gibbs paradox. Finally, the surface tension seems to have nothing to do with surface excess at all (Figure 2 in the previous paper7). It was shown that the slope of the decrease in surface tension against the logarithm of sodium dodecyl sulfate (SDS) concentration in a polyelectrolyte solution is 3.4 times as steep as that of just an SDS solution. The rationalization of this decrease in the surface tension of a cationic polymer solution can be briefly made as follows. The steep decrease from 60 to 40 mN m-1 takes place near the neutralization end point of the polymer cations with DS- anions. This decrease results from the destruction of the steric structure of the upper water molecular layers due to the floating-up of the neutralized polymers through the water layers. The floating-up due to buoyancy comes from the lower density of the neutralized polymers relative to that of water. This decrease is immediately followed by a steep increase in the surface tension to more than 60 mN m-1, where the neutralized polymers start to locate upon the surface despite the negative adsorption according to the Gibbs surface excess. Furthermore, the rather (7) Lee, J.; Moroi, Y. J. Colloid Interface Sci. 2004, 273, 645.
10.1021/la062014+ CCC: $33.50 © 2006 American Chemical Society Published on Web 11/17/2006
11206 Langmuir, Vol. 22, No. 26, 2006
Humphry-Baker et al.
Figure 2. Pyrene fluorescence spectra from the upper meniscus region (a) and the bulk (b) for just water saturated by pyrene. Figure 1. An optical cuvette to observe the fluorescence spectra of pyrene molecules in the upper meniscus region (a) and in the bulk (b). Table 1. Change in the Value of the I1/I3 Ratio with SDS Concentration I1/I3 ratio
SDS concn mmol dm-3
bulk
meniscus
0 0.80 1.00 1.33 2.00 4.00 8.00 16.0
1.77 1.74 1.71 1.66 1.69 1.62 1.21 1.15
1.74 1.70 1.69 1.60 1.56 1.62 1.28 1.14
simple conventional Gibbs adsorption just at the air/solution interface cannot explain the quite complex change in electric surface potential with SDS concentration.5 On the basis of this experimental evidence, the conventional Gibbs surface adsorption of soluble amphiphiles is quite difficult to accept. However, the new concept of a surface excess mentioned above is consistent with the several interfacial phenomena observed in soluble surfactant solutions. Experimental Section Pyrene was the same as that used in a previous paper, where the purity was verified by elemental analysis.8 SDS was purified by repeated recrystallization. The observed and calculated values (in parentheses) for the elemental analysis were in satisfactory agreement by weight percentage: C 49.88 (49.98), H 8.69 (8.74%). In addition, there was no minimum along surface tension versus concentration curve up to twice the critical micelle concentration (cmc), and the surface tension at 8.0 mM just below the cmc was 39.3 ( 0.1 mN m-1 at 298.2 K. The water used was distilled once after deionization by ion-exchange resin (Millipore, Milli-Q, resistivity 18 MΩ cm). SDS solutions were prepared using the water saturated by pyrene, where the pyrene concentration was 3.2 × 10-7 mol dm-3. A 1.5 mL portion of SDS solution was pipetted into a fluorescence optical cuvette of 1 × 1 cm. The excitation light beam of 4 mm height was adjusted to pass through the solution bulk or the upper meniscus region by moving the cuvette up or down (Figure 1). The fluorescence spectrophotometer was a Spex Fluorolog 112, and the excitation wavelength was 350 nm. The temperature was measured at 25 °C.
Results and Discussion In Table 1 are summarized the I1/I3 values of pyrene fluorescence spectra for the bulk solution and the upper meniscus region. The values for both the bulk and the meniscus are constant within experimental error, which indicates that the fluorescence (8) Yoshida, N.; Moroi, Y.; Humphry-Baker, R.; Gra¨tzel, M. J. Phys. Chem. A 2002, 106, 3991.
Figure 3. Pyrene fluorescence spectra from the upper meniscus region (a) and the bulk (b) for 2.0 mmol dm-3 SDS solution.
is from pyrene molecules mainly in the bulk. The I1/I3 values remain almost constant at around 1.7 in the aqueous bulk up to the surfactant cmc, although a very small decrease in this value can be seen. This decrease indicates that the hydrophobic interaction between pyrene and DS- ions increases with increasing concentration, as the mass-action model would suggest even at a concentration well below the cmc.9,10 A sharp decrease takes place around the cmc of SDS, 8.3 × 10-3 mol dm-3 at 298 K,11 due to the incorporation of pyrene molecules into the micellar interior.12 Figure 2 shows the pyrene fluorescence spectra of the bulk and the meniscus of the water only saturated with pyrene. As is clear from the difference in fluorescence intensity at 480 nm, the excimer can be observed from the enhanced concentration of pyrene molecules at the air/solution interface region just like ethanol.13,14 In other words, the pyrene molecule is surface active. Figure 3 illustrates the fluorescence spectra of pyrene in a 2.0 mmol dm-3 SDS solution. A large amount of the fluorescence from the excimer can be observed even at cmc/4 of SDS. This means that the probability for pyrene molecules to associate in the surface region is fairly high at such low SDS concentrations. In Figure 4, the spectra in an 8 mmol dm-3 SDS solution are shown, where both of the two spectra are superimposable, and no meniscus function can be seen. The extent of pyrene excimer was evaluated by the fluorescence intensity ratio of the spectral peaks at 373 (I1) and 480 nm (excimer, Iex), Iex/I1. The changes of this ratio with SDS concentration are plotted in Figure 5. The ratios from the meniscus are larger than those from the bulk at SDS concentrations up to (9) Morisue, T.; Moroi, Y.; Shibata, O. J. Phys. Chem. 1994, 98, 12995. (10) Moroi, Y.; Mitsunobu, K.; Morisue, T.; Kadobayashi, Y.; Sakai, M. J. Phys. Chem. 1995, 99, 2372. (11) The Chemical Society of Japan. Kagaku Benran, 5th ed.; Iwasawa, Y., Ed.; Maruzen: Tokyo, 2003. (12) Moroi, Y. Micelles: Theoretical and Applied Aspects; Plenum Press: 1992; Chapter 12. (13) Wilson, M. A.; Pohorille, A. J. Phys. Chem. B 1997, 101, 3130. (14) Kirpalani, D. M.; Toll, F. J. Chem. Phys. 2002, 117, 3874.
Pyrene Fluorescence at the Air/SDS Interface
Figure 4. Pyrene fluorescence spectra from the upper meniscus region (a) and the bulk (b) for 8.0 mmol dm-3 SDS solution.
Figure 5. Changes in the I480/I1 value with SDS concentration: 2 for the meniscus and O for the bulk.
the cmc, which means that the chance for pyrene molecules to collide with each other is higher in the surface region than that in the bulk. In addition, the fluorescence intensity from the excimer is more sensitive than that from the monomer, because there is no difference in the I1/I3 ratio between the bulk and the meniscus. The Iex/I1 ratios from the bulk gradually increase to the cmc, although there is one exception at 1.33 mmol dm-3, approximately cmc/6. This increase suggests an increase in the probability for pyrene molecules to associate, which is mediated by an increasing concentration of DS- ions. At a concentration of twice the cmc, however, the micellar concentration is much larger than the pyrene concentration and, therefore, the low occupation number of micelles by pyrene results in a low probability of an encounter with another pyrene, leading to a small ratio of 0.06. The value of the Iex/I1 ratio from the meniscus almost monotonically increases to the concentration of cmc/4 and then starts to decrease. It is quite certain that DS- ions concentrate at the surface region as a function of its bulk concentration and, therefore, pyrene molecules also concentrate in the region with increasing DS- ions, which results in an increased value of the ratio. However, the ratio values above cmc/4 decrease with increasing DS- concentration, contrary to the expectation of the conventional Gibbs surface excess. This decrease cannot be elucidated from the conventional concept of the surface excess just at the air/solution interface, where the surface excess remains
Langmuir, Vol. 22, No. 26, 2006 11207
almost constant over the surfactant concentration from cmc/2 to the cmc with a linear decrease in the plots of surface tension against the logarithm of concentration.15 If the surface excess or the adsorbed film of DS- ions truly locates at the air/solution interface, the pyrene concentration at the surface should remain almost constant with increasing DS- concentration from cmc/2 to the cmc, and therefore, the ratio value is supposed to stay constant with the increasing concentration up to the cmc, too. Unfortunately, the experimental observation is contrary to expectation. It is quite natural, therefore, to recognize that some molecular aggregates at the surface region forced the pyrene molecules to sit separately and decreased the chance for the pyrene molecules to encounter. The aggregates function just like micelles in the bulk, as mentioned below. In previous papers,2,4,5 the bimolecular layer like a cell membrane at some distance beneath the air/solution interface, say more than 6 nm, was suggested to rationalize other interfacial properties of surfactant solutions, which are an exchange rate constant of surfactant molecules between the bulk and the micelles16 and the rate constant of solubilizate molecules between the bulk and micelles.8,17 Such a bimolecular layer should solubilize the pyrene molecules and keep them located in it, which reduces the chance for the pyrene molecules to associate with each other. The experimental evidence presented here also supports this new description, instead of the conventional idea, for the surface excess. The bilayer aggregates disappear above the cmc, because the aggregates at the surface region are less stable energetically compared with globular micelles in the bulk above the cmc. Even at a concentration of 8.0 mmol dm-3 below the cmc, the micellar concentration is larger than the pyrene concentration, and there are enough micelles to accommodate all of pyrene molecules in the bulk due to the very large solubilization constant.9,10 In other words, all the pyrene molecules in the surface region are solubilized in the micelles in the bulk, and this results in no observation of pyrene molecules at the surface region, which gives rise to no difference in the ratio of the value between the bulk and the meniscus. The above fact also applies to the case for twice the cmc, where the chance for pyrene molecules to associate becomes much less, as is supported by the muchdiminished value of this ratio. What should be mentioned last is that the present work does not deny but accepts the Gibbs surface excess of SDS, because the surface excess was confirmed by an isotope study.18,19 The important point here is that the surface excess does not locate just at the air/water interface like in the conventional illustration, as is often drawn in many textbooks. LA062014+ (15) Mukerjee, P.; Mysels, K. J. Critical Micelle Concentrations of Aqueous Surfactant Systems; NSRDS-NBS 36, 1971; pp 11-18. (16) Zana, R. Surfactant Solutions: New Methods of InVestigation; Surfactant Science Series; Marcel Dekker: New York and Basel, 1987; Vol. 22, Chapter 8. (17) Zana, R. Surfactant Solutions: New Methods of InVestigation; Surfactant Science Series; Marcel Dekker: New York and Basel, 1987; Vol. 22, Chapter 5. (18) Muramatsu, M.; Tajima, K.; Sasaki, T. Bull. Chem. Soc. Jpn. 1968, 41, 1279. (19) Tajima, K.; Muramatsu, M.; Sasaki, T. Bull. Chem. Soc. Jpn. 1970, 43, 1991.
