1120
Langmuir 2004, 20, 1120-1125
Light-Sensitive Microemulsions Julian Eastoe,* Margarita Sanchez Dominguez, Hannah Cumber, and Paul Wyatt* School of Chemistry, University of Bristol, Bristol, BS8 1TS United Kingdom
Richard K. Heenan ISIS-CLRC, Rutherford Appleton Laboratory, Chilton, OXON, OX11 0QX United Kingdom Received November 4, 2003. In Final Form: November 28, 2003 A photodestructible surfactant, sodium 4-hexylphenylazosulfonate (C6-PAS), has been introduced to AOT-stabilized water-in-heptane microemulsions. Proton NMR spectra show that C6-PAS undergoes UVinduced decomposition, to yield a mixture of 4-hexylphenol and hexylbenzene. The photostationary state was determined by 1H NMR, indicating that nearly 90% of the initial photosurfactant had been destroyed, yielding non-surface-active hexylbenzene as the main product. This phototriggered breakdown gives rise to changes in adsorption and aggregation properties of C6-PAS, representing a novel route to induce microemulsion destabilization. When a series of microemulsions containing different amounts of C6-PAS were exposed to UV light, part of the dispersed water phase-separated. Small-angle neutron scattering (SANS) was used to follow the resulting UV-induced shrinkage of the water nanodroplets: a maximum volume decrease was found to be in the order of 60-70%. Kinetic SANS studies were also carried out in order to follow the changes in aggregation as a function of UV irradiation time. Multicontrast SANS experiments gave further insight; for example, it was demonstrated that the shell thickness remained constant. This study represents the first example of light-induced microemulsion destabilization.
Introduction The destabilization or resolution of microemulsions and emulsions is an important process in areas such as pharmaceuticals, organic synthesis, oil recovery, drug delivery, cosmetics, and nanotechnology. As such, these separations represent a product recovery/release mechanism. For emulsions, temperature- or electrolyte-induced approaches are common [e.g., refs 1 and 2]. More sophisticated methods include addition of demulsifiers and/or centrifugation,3,4 centrifugal pulse-voltage,5 nonuniform electric fields,6 magnetofluidized beds,7 electric and magnetic fields,8,9 high electric fields,10 and microwave-induced destabilization.11 On the other hand, owing to thermodynamic stability microemulsions represent a more difficult challenge and several methods have been used to affect phase separation of these systems: pervaporation,12 * To whom correspondence should be addressed. Tel: U.K. + 117 9289180. Fax: U.K. + 117 9250612. E-mail: julian.eastoe@ bristol.ac.uk. (1) Euston, S. R.; Finnigan, S. R.; Hirst, R. L. J. Agric. Food Chem. 2001, 49, 5576. (2) Villamagna, F.; Whitehead, M. A.; Chattopadhyay, A. K. J. Dispersion Sci. Technol. 1995, 16, 105. (3) Sullivan, A. P.; Kilpatrick, P. K. Ind. Eng. Chem. Res. 2002, 41, 3389. (4) Radvinskii, M. B.; Kikhteva, V. I. Vodosnabzh. Sanit. Tekh. 1973, 10, 16. (5) Li, S.; Mo, R.; Zhang, H.; Yan, Z. Mo Kexue Yu Jishu 1994, 14, 11. (6) Filina, R. A.; Petrov, A. A.; Prishchenko, N. P. Tekh. Nauki 1975, 3, 11. (7) Beril, I. I.; Bologa, M. K.; Kozhukhar, I. A. Elektron. Obrab. Mater. 1978, 6, 34. (8) Goncalves de Oliveira, R. C.; Figueiredo, A. M. P.; Monteiro de Carvalho, C. H. Braz. Pedido PI 1998, 19. (9) Li, K. Mo Kexue Yu Jishu 1996, 16, 50. (10) Gang, L.; QiongHua, L.; PanSheng, L. J. Membr. Sci. 1997, 128, 1. (11) Maue, R.; Moll, F. Acta Pharm. Technol. 1987, 33, 225. (12) Aouak, T.; Moulay, S.; Hadj-Ziane, A. J. Membr. Sci. 2000, 173, 149.
addition of electrolytes,13-15 low or high temperature, pH change and saturation with water,15 and finally addition of antisolvents.16 In the case of pervaporation, mechanical removal of one of the components promotes destabilization, whereas the solvent addition method induces desorption of the stabilizing surfactant layer. Other techniques generally involve a shift in hydrophobic/hydrophilic interactions of surfactants: for example, addition of electrolytes (ionic surfactants), temperature jumps (nonionics), and pH change (zwitterionics). In this paper, a novel approach to microemulsion destabilization is reported: if a photodestructible surfactant is added to a typical microemulsion formulation (heptane/AOT/water), selective photochemical breakdown of this component can be induced by UV light, hence shifting the location of the sample in the phase diagram. This paper expands on the initial report of photoresponsive microemulsions.17 Here the photochemical breakdown followed by 1H NMR and multiple contrast and kinetic data are presented; hence a more comprehensive account of these unusual systems is offered. The photosurfactant used in this investigation, sodium 4-hexylphenylazosulfonate (C6-PAS, shown in Scheme 1), has been employed previously in emulsion polymerization,18,19 for which a successful separation of poly(methyl methacrylate) (PMMA) latex particles was (13) Keiser, B. A.; Varie, D.; Barden, R. E.; Holt, S. L. J. Phys. Chem. 1979, 83, 1276. (14) Desnoyers, J. E.; Quirion, F.; Hetu, D.; Perron, G. Can. J. Chem. Eng. 1983, 61, 672. (15) Hazbun, E. A.; Schon, S. G. U.S. Patent Application US4770670, 1988. (16) Ingelsten, H. H.; Beziat, J. C.; Bergkvist, K.; Palmqvist, A.; Skoglundh, M.; Hu, Q.; Falk, L. K. L.; Holmberg, K. Langmuir 2002, 18, 1811. (17) Eastoe, J.; Sanchez-Dominguez, M.; Cumber, H.; Burnett, G.; Wyatt, P.; Heenan, R. K. Langmuir 2003, 19, 6579. (18) Mezger, T.; Nuyken, O.; Meindl, K.; Wokaun, A. Prog. Org. Coat. 1996, 29, 147. (19) Dunkin, I. R.; Gittinger, A.; Sherrington, D. C.; Wittaker, P. J. Chem. Soc., Perkin Trans. 2 1996, 1837.
10.1021/la0360761 CCC: $27.50 © 2004 American Chemical Society Published on Web 01/13/2004
Light-Sensitive Microemulsions Scheme 1. Photolysis of Sodium 4-Hexylphenylazosulfonate (C6-PAS)
demonstrated based on the action of incident UV. In AOTstabilized water-in-oil (w/o) microemulsions, isothermal UV-triggered percolation has been induced by employing trace levels of light-sensitive solubilizates such as 9-hydroxymethylanthracene.20,21 For model emulsions, comprising azobenzene-containing polyelectrolytes, UVvisible light has been used to control emulsion type (oil or water continuous medium) in the presence of NaNO3.22 Recently light-induced changes in adsorption and aggregation for a stilbene-containing gemini photosurfactant in aqueous solutions have been reported.23 Here the utility of photosurfactants such as C6-PAS for changing properties of AOT-stabilized water-in-oil microemulsions is explored, therefore extending the applications of such light-sensitive stabilizers to oil-water interfaces and microemulsion droplets, which are systems of interest to the nanotechnologist. Experimental Section Materials. Photosurfactant 1 (sodium 4-hexylphenylazosulfonate, C6-PAS) was synthesized as described elsewhere.18,19 Melting point, 159-162 °C; literature, 155-157 °C.19 Elemental analysis found: C, 49.51; H, 5.68; N, 9.60; S, 10.71%. Required for C12H17N2 O3SNa: C, 49.30; H, 5.90; N, 9.6; S, 11.00%. δH (CD3OD): 7.84 (2H, d, ArH), 7.40 (2H, d, ArH), 2.72 (2H, t, ArCH2CH2), 1.67 (2H, m, ArCH2CH2), 1.36 (6H, m, 3 × CH2), and 0.90 (3H, t, CH3). νmax/cm-1 (KBr): 3054w, 2959m, 2927s, 2872m, 2854s, 1605w, 1513m, 1466m, 1419w, 1284m, 1235s, 1159m, 1067s, 1016w, 954w, 897w, 846m, 833m, 760m, 722w, and 648s. The Krafft temperature (Tk ) 25 °C) was determined by electrical conductivity, and the critical micelle concentration (cmc) by surface tension using a Kruss K10 Du Nouy Ring instrument (cmc at 30 °C ) 2.25 × 10-2 mol dm-3). These measured properties compare favorably with literature values.18,19 The water-in-oil microemulsions were prepared containing AOT sodium salt (Sigma, 99% min, purified from CHCl3 and activated charcoal before use), n-heptane (Aldrich, HPLC grade, 99% min), d-heptane (Aldrich, 99%), D2O (Fluorochem, 99.9%), and 1. The surfactant is hence a mixture of C6-PAS and AOT with a composition defined by X1(%) ) 100{[1]/([AOT] + [1])}; the total concentration of surfactant {[1] + [AOT]} was either 0.05 or 0.10 mol dm-3. In the initial dispersion ratio prior to UV irradiation, w ([water]/[surfactant]) was kept constant at w ) 40. Techniques. Glassware was thoroughly cleaned as previously described.24 1H NMR spectra were measured with a JEOL Delta GX/270, for microemulsion samples at X1(%) ) 7.5% and total surfactant concentration {[1] + [AOT]} ) 0.05 mol dm-3. Changes in the spectra were followed as a function of irradiation time. For (20) Wollf, T.; Hegewald, H. Colloids Surf., A 2000, 164, 279. (21) Ness, D.; Cichos, U.; Wolff, T. Ber. Bunsen-Ges. Phys. Chem. Chem. Phys. 1996, 100, 1372. (22) Porcar, I.; Perrin, P.; Tribet, C. Langmuir 2001, 17, 6905. (23) Eastoe, J.; Sanchez-Dominguez, M.; Wyatt, P.; Beeby, A.; Heenan, R. K. Langmuir 2002, 18, 7837. (24) Downer, A.; Eastoe, J.; Pitt, A. R.; Penfold, J.; Heenan, R. K. Colloids Surf., A 1999, 156, 33.
Langmuir, Vol. 20, No. 4, 2004 1121 reasons of economy, the microemulsions for NMR analysis were prepared in d-cyclohexane (Goss Scientific, 99.5% D) instead of d-heptane (note h-heptane and d-heptane were used for droplet sizing by neutron scattering). This switch in solvent is known to affect phase diagram and maximum water uptake; hence, the compositions of the cyclohexane samples were arranged to provide single stable phases at 25 °C. The use of cyclohexane in place of heptane for these NMR experiments will have no significant effect on the conclusions concerning photoproduct formation. Reference microemulsions containing standards (Aldrich, 99%) 4-octylphenol (structurally similar to 2) and 4-octylbenzene (similar to 3) were also characterized in order to determine the spectra of the expected photoproducts. The octyl- instead of the hexyl-substituted product had to be used since the latter is not available commercially; nevertheless, the 1H NMR spectra and chemical shifts are expected to be very similar (octyl vs hexyl). Photoexperiments. Samples were irradiated with a 100 W highpressure Hg lamp (LPS-220/250 power supply from PTI, lamp housing from HI-TECH Scientific, and Mercury short arc photo optic lamp HBO 103 W/2), which was polychromatic to maximize the photon flux. The samples for the first series of small-angle neutron scattering (SANS) experiments (function of X1(%)) were stirred with a small magnetic flea; samples for SANS were taken after the irradiated samples were allowed to equilibrate at an appropriate temperature (25 °C for X1(%) ) 5-10%, 40 °C for X1(%) ) 0-2.5%). For the kinetic experiments, the samples were irradiated in quartz Hellma cells (1 mm) and they were left to equilibrate (phase-separate) in the sample changer used for neutron scattering and therefore they were not stirred. Small-Angle Neutron Scattering. SANS experiments were carried out in the time-of-flight LOQ instrument at ISIS, U.K., where incident wavelengths are 2.2 e λ e 10 Å, resulting in an effective Q range of 0.009 f 0.249 Å-1, or at the D22 diffractometer at ILL (Grenoble, France) using a neutron wavelength of λ ) 10 Å at two different detector distances to cover a Q range of 0.0024 f 0.37 Å-1. Absolute intensities for I(Q) (cm-1) were determined to within (5% by measuring the incoherent scattering from 1 mm of H2O at ILL, whereas at ISIS a partially deuterated polymer standard was employed.25 D2O water droplets were contrasted against h-surfactant and h-heptane (core contrast), whereas h-surfactant shell was contrasted against D2O and d-heptane (shell contrast). Measurements were conducted at 25 °C except for X1 ) 0 and 2.5% which were done at 40 °C. Rectangular Hellma quartz cells were of 1 mm path length for core contrast and 2 mm for shell contrast. The neutron beam was scattered from the upper transparent microemulsion phase, missing the denser separated part in the case of irradiated samples. The experiments were divided in four series. First, equilibrium experiments (core contrast, ILL) were performed in order to study the effect of the concentration of 1, measuring the scattering before and after UV irradiation for a fixed time (1 h for {[1] + [AOT]} ) 0.05, 2 h for {[1] + [AOT]} ) 0.10); the completion of the reaction was confirmed by UV-vis absorption. Equilibrium experiments at a chosen concentration (X1(%) ) 7.5% and {[1] + [AOT]} ) 0.05 mol dm-3) and two different contrasts (core and shell, ILL) were also performed. The third set of experiments consisted of on-line kinetic runs carried out to follow changes as a function of irradiation time with the same kinds of samples as mentioned above with both contrasts (ISIS). These were irradiated (100 W Hg lamp) for fixed intervals in the SANS sample changer, away from the neutron beam, and SANS was run immediately afterward. The irradiation intervals were gradually increased until a stationary state was reached, after about 10 min. Finally, experiments were carried out in order to investigate the effects of different photoproducts on the microemulsion stability and structure (core contrast, ILL). For this purpose, a microemulsion was prepared simulating the composition at X(1+PP)(%) ) 7.5% and {[1] + [AOT] + [photoproducts]} ) 0.05 mol dm-3 at the end of the photolysis (sample ROP1). Hence this system contained AOT, 1, 4-octylphenol, and 4-octylbenzene at concentrations at the photostationary state as determined by 1H NMR. Note that for the same reasons as for the 1H NMR experiments, octyl- instead of hexyl-substituted products were (25) Summers, M.; Eastoe, J.; Davis, S.; Du, Z.; Richardson, R. M.; Heenan, R. K.; Steytler, D. C.; Grillo, I. Langmuir 2001, 17, 5388.
1122
Langmuir, Vol. 20, No. 4, 2004
Eastoe et al.
Figure 1. Photoinduced destabilization of AOT/water/heptane microemulsions containing C6-PAS. used. Additional samples (ROP2-ROP7) were done in a similar way except that one or two of the photoproducts were removed as appropriate. SANS data were fitted using the interactive FISH program,26 which is a flexible multimodel package comprising a variety of different form factors P(Q), structure factors S(Q), and polydispersity functions. After extensive trials and based on previous literature,27 the most appropriate model was found to be for Schultz polydisperse two-shell hard spheres. Whenever necessary, an attractive structure factor was also tested in order to account for weak interactions; nevertheless the structural information obtained with both methods was very similar. Porod and Guinier approximations were also employed, and full accounts of the relevant scattering laws are given elsewhere.28 In the case of multicontrast experiments performed at ILL (equilibrium experiments), a simultaneous model was used to analyze the core and shell contrast experiments at the same time in order to obtain the more realistic representation of the system [e.g., refs 26 and 29]. This was not possible in the case of the kinetic runs since different irradiation times were employed for different contrasts. The Schultz polydispersity factor was a fitting parameter and in all cases was between 0.20 and 0.23, which is consistent with previous studies of AOT microemulsion systems [e.g., ref 29].
Results and Discussion (a) Photochemistry and NMR Studies. Figure 1 shows the destabilization seen for these microemulsion systems upon irradiation with UV light: initially the sample is transparent (a) since nanometer-sized aqueous droplets are present. After a burst of UV light (b), a milky emulsion containing micron-sized droplets begins to form, which eventually resolves into a clear upper microemulsion phase and a denser, opaque water-surfactant mixture (c). It is known that light causes an irreversible photochemical change for such azosulfonates.18,19 These previous studies indicate that the photolysis pathway is strongly dependent on solvent type or microenvironment. Therefore, as depicted in Scheme 1, an ionic decay mechanism would be favored in aqueous solution, whereas in the more hydrophobic reverse micellar or w/o environment, a radical mechanism may dominate. The two different decay (26) Heenan, R. K. FISH Data Analysis Program; Rutherford Appleton Laboratory; Report RAL-89-129; CCLRC: Didcot, U.K., 1989. (27) Bumajdad, A.; Eastoe, J.; Griffiths, P.; Steytler, D. C.; Heenan, R. K.; Lu, J. R.; Timmins, P. Langmuir 1999, 15, 5271. (28) Zemb, T.; Lindner, P. Neutron, X-Ray and Light Scattering: Introduction to an Investigative Tool for Colloidal and Polymeric Systems; Elsevier Science Publishers B.V.: Amsterdam, 1991; pp 11 and 199. (29) Eastoe, J.; Hetherington, K. J.; Sharpe, D.; Dong, J.; Heenan, R. K.; Steytler, D. C. Colloids Surf., A 1997, 128, 209.
Figure 2. Partial 1H NMR spectra of AOT/water/d-cyclohexane microemulsions containing (a) 4-octylphenol, (b) octylbenzene, (c) and nonirradiated C6-PAS; curves d-f are samples from curve c recorded after different irradiation times.
pathways would yield photoproduct 2 (4-hexylphenol) or 3 (hexylbenzene), respectively. Figure 2a,b shows 1H NMR spectra in the high δ aromatic region for microemulsions containing only AOT as the surfactant and related amounts (X(%) ) 7.5%) of standards 4-octylphenol (equivalent to spectrum for 2) and octylbenzene (equivalent to spectrum for 3), respectively. Figure 2c-f shows corresponding spectra for microemulsions stabilized by the C6PAS + AOT mixture (X1(%) ) 7.5%, {[AOT] + [1]} ) 0.05 mol dm-3) as a function of irradiation time. It is clear that the peaks from surfactant 1 (δ ) 7.86 and 7.26) decay as photolysis proceeds, whereas new peaks also appear. The lines at δ ) 6.71 and δ ) 6.88 clearly indicate the presence of an alkylphenol 2, while the multiplet at δ ) 7.1 is characteristic of an alkylbenzene 3. The temporary doublets that appear at intermediate irradiation times (Figure 2d,e) around δ ) 8.60 and 7.55 could be attributed to an intermediate diazonium salt.30,31 This is supported by the fact that these lines reduce with irradiation and that they are essentially absent in the final spectrum. Analyses of these NMR spectra suggest that the reaction is essentially complete, and a photostationary state is reached after about 110 min of irradiation, corresponding to a composition of 10.6% of 1 (or an unknown photoproduct, see below), 37.0% of 2, and 52.4% of 3, based on the initial amount of 1. In terms of remaining 1, it is difficult to discriminate between this and other potential (hard to identify) photoproducts, since the doublet at δ ) 7.26 indicating the presence of 1 overlaps partially with the multiplet from 3. It is possible that as the reaction proceeds and the sample goes from monophasic to Winsor (30) Adenier, A.; Bernard, M.-C.; Chehimi, M. M.; Cabet-Deliry, E.; Desbat, B.; Fagebaume, O.; Pinson, J.; Podvorica, F. J. Am. Chem. Soc. 2001, 123, 4541. (31) Suzuki, H.; Nonoyama, N. Tetrahedron Lett. 1998, 39, 4533.
Light-Sensitive Microemulsions
Langmuir, Vol. 20, No. 4, 2004 1123
Table 1. Fitted Droplet Core Radius as a Function of X1(%) and {[AOT] + [1]}, before and after Irradiation (Core Contrast) core radius, core radius, {[AOT] + [1]}/ before after % volume mol dm-3 X1(%) irradiation/Å irradiation/Å shrinkage 0.05 0.05 0.05 0.05 0.05 0.10 0.10 0.10 0.10 0.10
0.0 2.5 5.0 7.5 10.0 0.0 2.5 5.0 7.5 10.0
55.2 56.9 58.4 58.1 58.8 52.5 53.0 55.0 55.7 57.5
55.1 45.1 43.2 43.1 42.9 51.6 44.1 42.6 37.4 35.8
0.3 50.2 59.7 59.1 61.0 5.1 42.4 53.7 69.7 75.9
II, the position of the peaks could shift slightly (peak at δ ∼ 7.8). Nevertheless, it has been shown that the main photoproduct is non-surface-active alkylbenzene 3, and this indicates that the reaction is preferentially following a radical mechanism, as expected. Considering the final composition (total surfactant concentration decreases to ∼0.045 mol dm-3), the effective w value increases from w ) 40 to about w ) 44; furthermore, the major photoproduct (hexylbenzene) adds to the continuous oil phase. (b) Small-Angle Neutron Scattering. Equilibrium Experiments as a Function of X1(%), Core Contrast. A series of microemulsions from X1 ) 0 to X1 ) 10% and {[AOT] + [1]} ) 0.05 and 0.10 mol dm-3 were prepared as described in the Experimental Section. The phase separation shown in Figure 1 could be related to shrinkage of the D2O droplet cores, due to the evident loss of water and therefore reduction in effective w. Table 1 is a summary of SANS analyses of samples as a function of X1, which were irradiated for sufficiently long to achieve the photostationary state: the mean droplet core radius becomes smaller after irradiation in all cases, consistent with the accompanying loss of water from the dispersion. Furthermore, in all cases the percent of volume shrinkage (related to the volume of lost water) is around 50-60% for samples when {[AOT] + [1]} ) 0.05 mol dm-3 and ∼4076% with {[AOT] + [1]} ) 0.10 mol dm-3. This dramatic shrinkage is evidently due to the destruction of nearly all present C6-PAS as confirmed by 1H NMR. Hence, the change in chemical nature of the surfactant moves the system to a different position in the phase diagram and therefore it becomes unstable and phase-separates. The fact that there is a significant effect for a sample containing as little as 2.5% of 1 can be explained in terms of a different cloud point. For a sample containing 2.5% of 1, this temperature is ∼30 °C, whereas the samples containing 5-10% were clear at 25 °C; the effect of irradiation in the first sample is pronounced since it is positioned near a Winsor II microemulsion phase boundary. These concepts are enlarged in ref 27. On the other hand, samples with higher X1 are further from the phase boundary, but destruction of nearly 5-10% of the total surfactant is enough to give a similar effect. The results also generally show that as X1 is increased, the radius of the particles slightly increases. Increasing X1 represents replacement of the double-chain AOT, which favors negative curvature, with 1, which is a single-chain surfactant and should favor positive curvature, and as the curvature becomes more positive, the water solubilization of the system (wmax) is increased. Phase behavior studies confirmed this, since at 25 °C the maximum water solubility (wmax) increased from 29 for X1(%) ) 0 to 38, 88, 73, and 52 for X1(%) ) 2.5, 5, 7.5, and 10%, respectively. The first two samples crossed the low temperature or solubilization (Winsor II) bound-
Figure 3. SANS data from AOT/water/heptane containing C6PAS at X1(%) ) 7.5 and {[AOT] + [1]} ) 0.05 M with core (b) and shell (O) contrasts before (a) and after (b) irradiation. Lines are simultaneous fits to both data sets.
ary, and last two systems phase-separated at a haze boundary.32 The wmax value for X1(%) ) 5 is around the middle of the two boundaries; this behavior is observed since when X1(%) is increased the whole single-phase region is shifted to lower temperatures. The increase in wmax when 1 is present compares favorably with previous studies of mixtures of double-chain with single-chain surfactants in microemulsions.27 Equilibrium Experiments, Core and Shell Contrast. To obtain more detailed information on the interfacial structure and the effects of irradiation, equilibrium experiments with core and shell contrast were carried out. Shell contrast (d-water/h-surfactant/d-oil) is used to see a “hollow shell” of surfactant, with a characteristic interference pattern. As shown elsewhere, a robust analysis can be performed by a simultaneous fit to data from the three different contrasts (core, shell, drop).29,33,34 Although here only core and shell data are employed, this is still a more realistic representation than merely analyzing a single contrast data set. As shown in Figure 3, the water core shrinkage from core-shell simultaneous analysis of 63 Å to 40 Å is in good agreement with the core contrast analysis (see Table 1). Since in previous studies29 there was no evidence for significant oil penetration into the surfactant layer for AOT systems, the same situation was considered here. Fitted shell thickness t is given in Figure 3, and this dimension remains almost unaltered: the value of ∼9 Å (32) Eastoe, J.; Dong, J.; Hetherington, K. J.; Steytler, D. C.; Heenan, R. K. J. Chem. Soc., Faraday Trans. 1996, 92, 65. (33) Fletcher, P. D. I.; Howe, A. M.; Robinson, B. H. J. Chem. Soc., Faraday Trans. 1 1987, 83, 985. (34) Heenan, R. K.; Eastoe, J. J. Appl. Crystallogr. 2000, 33, 749.
1124
Langmuir, Vol. 20, No. 4, 2004
Eastoe et al.
Figure 4. SANS data from AOT/water/heptane containing C6PAS at X1(%) ) 7.5 and {[AOT] + [1]} ) 0.05 M with core (a) and shell (b) contrasts at different irradiation times: for core, (b) 0, (O) 3 min, (1) 5 min, and (3) 10 min; for shell, 0, 40, 50, and 60 s, respectively. Lines are fits from model fitting. Table 2. Fitted Droplet Core Radius as a Function of Irradiation Time for the Kinetic Run, Using Different Analysis Methods (Core Contrast) irradiation time/min
core radius, FISH model/Å
core radius, Porod/Å
core radius, Guinier/Å
0 3 5 10
58.0 56.3 49.2 38.2
66.7 62.7 56.2 42.2
78.8 73.0 64.1 50.7
is in good agreement with previous studies of AOTstabilized water-in-heptane microemulsions.29 This is not surprising since the molecular length of both surfactants is very similar, and therefore destruction of 1 in this context is essentially equivalent to reducing the effective surfactant concentration. Kinetic Experiments, Core and Shell Contrast. To further investigate the effectiveness of light-induced destabilization of these systems, on-line kinetic runs with core and shell contrasts were carried out. Figure 4 shows data for a water-in-heptane sample at X1 ) 7.5% and {[AOT] + [1]} ) 0.05 mol dm-3 at different stages with core (a) and shell (b) contrast; it can be seen that I(Q) decreases with irradiation time, consistent again with the loss of water from the cores and a related gradual shrinkage of water droplet core radius (Tables 2 and 3). It is important to note that the reaction conditions for these kinetic runs were not strictly the same as for the equilibrium experiments: in this case, no stirring was provided to the reaction since it was carried out in 1 mm path length cells. Nevertheless, similar values of core radii at equilibrium were obtained, and the values for both contrasts were in good agreement. Figure 5 shows the analyses by Guinier
Figure 5. Guinier (a) and Porod (b) approximations of data from Figure 4a. Symbols are as in Figure 4a. For panel a, lines are linear fits to the low Q region; for panel b, lines are guides to the eye. Table 3. Fitted Form Factor Droplet Core Radius as a Function of Irradiation Time for the Kinetic Run (Shell Contrast) irradiation time/s
core radius/Å
0 40 50 60
62.8 50.7 43.9 40.1
(a) and Porod (b) approximations for the core contrast experiment (assuming spherical particles). The Porod plot shows a gradual shift in Qmax to higher Q, and the reduction in the asymptotic intensity for Q > 0.10 Å-1 is consistent with a lower total surface area,28 which would be expected if surfactant 1 and water were eliminated from the dispersion. Likewise, the Guinier analysis shows a gradual reduction in both the gradient and intensity, consistent with a reduction of water volume fraction. For both methods, the analyses are consistent with droplet shrinkage and the values of core radius are within 10-30% in comparison with the results obtained from full model fitting (see Table 2); interactions (accounted for in the full critical scattering model) may explain any differences with the dimensions from model fitting. Role of Photoproduct, Core Contrast. To investigate how the photoproducts 2 and 3 affect the equilibrium stability and structure of the microemulsions, a series of samples (X(1+PP)(%) ) 7.5% and {[1] + [AOT] + [photoproducts]} ) 0.05) containing the different species were studied. These systems were based on the composition of the photostationary state obtained by 1H NMR. Table 4 summarizes the results from model fitting; all samples displayed Winsor II phase separation at 25 °C. Consider the results from nonirradiated (radius ∼ 58 Å) and irradiated (radius ∼ 43 Å) samples. Sample ROP1
Light-Sensitive Microemulsions
Langmuir, Vol. 20, No. 4, 2004 1125
Table 4. Fitted Form Factor Droplet Core Radius and Compositions for Samples Containing Different Amounts of Added Photoproducts sample
components
core radius/Å
ROP1 ROP2 ROP3 ROP4 ROP5 ROP6 ROP7
AOT + 1 + 4-octylphenol + 4-octylbenzene AOT AOT + 1 AOT + 1 + 4-octylphenol AOT + 1 + 4-octylbenzene AOT + 4-octylphenol AOT + 4-octylbenzene
49.2 52.0 52.9 50.2 52.5 47.8 51.8
represents a “simulated” sample at the photostationary state, containing both photoproducts and remaining photosurfactant 1. Although this sample (radius ∼ 49 Å) did not exactly reproduce the result obtained after irradiation mentioned above (43 Å), this indicates that the photoproducts affect the stability of the microemulsion. There are two main reasons for the difference in radius between simulated and irradiated samples: the first one arises from the use of octyl- instead of hexyl-substituted phenol and benzene. The second is, as discussed in the NMR section, that it is possible that 1 is totally photolyzed, and instead another unknown photoproduct is present, which destabilizes the microemulsion. Of the two main photoproducts, alkylphenol 2 has the greatest effect on the phase behavior/structure: the samples containing only this photoproduct (ROP4 and ROP6) have a smaller radius and the lowest intensity (Table 4). Other samples have similar radii, regardless of the presence of 1 and 3. The fact that the radius for ROP2 (containing only AOT) is also ∼52 Å, and importantly a Winsor II system, supports the phase behavior observations, since without the added C6-PAS at X1(%) ) 7.5% the sample cannot remain monophasic at w ) 44. These results demonstrate that the destabilization is due to a shift of the sample to a different position in the phase diagram and a more negative curvature in the absence of photosurfactant 1. However, another factor is the formation of 2, which could be acting as a cosurfactant, hence creating more available surface and affecting the curvature and bending rigidity of the system.35,36 (35) Szleifer, I.; Kramer, D.; Ben-Shaul, A.; Roux, D.; Gelbart, W. M. Phys. Rev. Lett. 1988, 60, 1966. (36) Farago, B.; Richter, D.; Huang, J. S.; Safran, S. A.; Milner, S. T. Phys. Rev. Lett. 1990, 65, 3348.
Summary and Conclusions A novel approach to microemulsion destabilization has been presented. The use of UV light to induce water separation represents a straightforward method, as there is no need for addition of extra components or modification of the conditions. 1 H NMR analysis shows clearly the destruction of ∼90100% of C6-PAS, and this goes in hand with the change in appearance of the system, which undergoes phase separation; the main photolysis pathway is via the radical mechanism to preferentially produce hydrocarbon 3, but phenol 2 is also formed. It has been demonstrated that alkylphenol 2 affects the stability of the microemulsions, possibly due to adsorption at the interface acting as a cosurfactant, whereas 3 has little effect and therefore it only adds to the continuous oil phase, diluting the microemulsion slightly. As shown by Table 1 and Figures 3-5, SANS studies show that this photolysis causes the shrinkage of water nanodroplets, and it was found that even with a small amount of C6-PAS (X1(%) ) 2.5%) a large light-induced shrinkage is observed, obtaining a maximum effect of 76% for X1(%) ) 10%. A simultaneous analysis of SANS core-shell contrast series has been used to investigate the effect of photolysis in the shell thickness, and it was found that this remains almost unaltered. The effect of light in this system as a function of time was also studied, and a gradual shrinkage was observed; therefore it could be possible to control the rate of the microemulsion destabilization as required by regulating the light intensity. It is hoped that this concept can be applied to areas such as the recovery of nanoparticles from microemulsions. Acknowledgment. M. S.-D. is grateful to the Mexican organization CONACYT (National Council of Science and Technology, Grant No. 151737) for a scholarship. The work was supported by a grant from the ACORN consortium (EPSRC-DTI). We also acknowledge CLRC for allocation of beam time at ISIS and grants toward consumables and travel, and the same acknowledgments are due to ILL. Special thanks go to our ILL local contact Isabelle Grillo. LA0360761
