pubs.acs.org/Langmuir © 2009 American Chemical Society
Reversible Switching of Electrical Conductivity in an AOT-Isooctane-Water Microemulsion via Photoisomerization of Azobenzene Markus Bufe and Thomas Wolff* :: Technische Universitat Dresden, Physikalische Chemie, D-01062 Dresden, Germany Received February 17, 2009. Revised Manuscript Received April 2, 2009 The electrical conductivity of microemulsions composed of aerosol OT (AOT), isooctane, and water as a function of temperature was studied in the absence and presence of azobenzene, and consequences of an in situ transcis photoisomerization of azobenzene were investigated. A conductivity onset upon raising the temperature of a waterin-oil microemulsion indicates percolation. Small amounts (0.1-5% w/w) of solubilized azobenzene induce higher percolation temperatures Tp (by up to 19 K), and photoisomerization of azobenzene shifts Tp back to values that may be below Tp in the absence of azobenzene. Consequently, the microemulsion can be switched from nonconducting to conducting by exposing samples to UV-light at λ > 310 nm, without varying temperature or composition. The effect reverts within several minutes after turning off the irradiation lamp through thermal reisomerization. By that, reversible switching of electrical conductivity is brought about.
1. Introduction Aerosol OT (AOT, i.e., sodium bis-2-ethylhexylsulfosuccinate, Scheme 1) is a frequently used surfactant that gives rise to a large variety of colloidal structures in fluid systems. Liquid crystalline1 and micellar solution structures2 in binary mixtures of AOT and water, glycerol, or formamide are known, as well as inverted structures in nonpolar solvents.3 Also, thermodynamically stable microemulsions may be formed in the presence of a third component, a nonpolar oil. Microemulsions composed of oil, water, and AOT may exist in the visually monophasic water-in-oil (w/o) or oil-in-water (o/w) structure as well as in the intermediate bicontinuous (spongelike) structure, depending on the water-tooil ratio, on the kind of oil, on the AOT content, and on temperature and pressure. Corresponding phase diagrams can be found in textbooks, for example, ref 4. A pseudobinary phase diagram of an AOT microemulsion system (Figure 1, derived from neutron scattering experiments5) reveals that transitions between the w/o and the bicontinuous structure (and vice versa) are possible by passing the percolation line (shown in Figure 1 at 60-90% n-decane) in that either the composition or the temperature is changed. This transition can be monitored by conductivity measurements, as the bicontinuous structure is electrically conductive due to the fact that channels of water and oil separated by a flexible surfactant layer exist; that is, the water phase is continuous and extends from one end of the system to the other, in contrast to the w/o structure. An example of a conductivity versus temperature curve for a pure microemulsion (without solubilizate) is given in
Figure 2 (0). The onset of conductivity upon heating indicates the transition, which is also called percolation. At higher temperatures (>35 °C), the curve represents features of entering the upper two-phase region, cf. Figure 1. Percolation temperatures (Tp, at fixed composition in similar AOT systems) can be influenced by electric fields6 or by the addition of small amounts of certain compounds that affect Tp in a specific way.7 We have found that percolation can be induced;at constant temperature and composition;when small amounts of photochemically reactive compounds are solubilized and irradiated, provided photoeduct and photoproduct induce different percolation temperatures. In this respect, the influence of neutral and ionic anthracene derivatives, of N-methyl-2-quinolone, and of their photodimerization was studied8 as well as the effects of hydroxystilbazolium cations and their trans-cis photoisomerizations;9 see N-methylhydroxystilbazolium bromide (HSB) as an example in Figure 2. By exposing samples to light of suitable wavelengths, isothermal conductivity changes (either turning on or off) at fixed composition were obtained in all the above cases, whose amplitudes can amount up to 3 mS cm-1. In this paper, we present the first truly reversible example of photochemically induced switching of conductivity. In principle, reversibility in such systems can be brought about in two ways: (i) in that the photochemical reaction inducing the conductivity switch can be photochemically reverted by reirradiation at a different (mostly lower) wavelength or (ii) in that a photoreaction is used that is thermally reversible. Previous attempts in this respect were unsatisfactory for practical application. An example for (i) is the photochemical reversion of the photodimerization
*To whom correspondence should be addressed. E-mail: thomas.
[email protected]. Telephone: +49-351-46333633. Fax: +49351-46333391.
(6) Schlicht, L.; Spilgies, J.-H.; Runge, F.; Lipgens, S.; Boye, S.; Schulbel, D.; Ilgenfritz, G. Biophys. Chem. 1996, 58, 39–52. (7) (a) Garcia-Rio, L.; Leis, J. R.; Mejuto, J.-C.; Pena, E.; Iglesias, E. Langmuir 1994, 10, 1676–1683. (b) Dasilva-Carvalhal, J.; Garcia-Rio, L.; Gomez-Diaz, D.; Mejuto, J.-C.; Rodriguez-Dafonte, P. Langmuir 2003, 19, 5975–5983. (c) Chakraborty, I; Moulik, S. P. J. Colloid Interface Sci. 2005, 289, 530–541. (d) Mehta, S. K.; Sharma, S. J. Colloid Interface Sci. 2006, 296, 690–699. (8) (a) Nees, D.; Cichos, U.; Wolff, T. Ber. Bunsen Ges. 1996, 100, 1372–1373. (b) Wolff, T.; Hegewald, H. Colloids Surf., A 2000, 164, 279–285. (c) Bufe, M.; Wolff, T. Phys. Chem. Chem. Phys. 2006, 8, 4222–4227. (9) Wolff, T.; Nees, D. Prog. Colloid Polym. Sci. 1998, 111, 113–116.
(1) (a) Nees, D.; Wolff, T. Langmuir 1996, 12, 4960–4965. (b) Nees, D.; Blenkle, M.; Koschade, A.; Wolff, T.; Baglioni, P.; Dei, L. Prog. Colloid Polym. Sci. 1996, 101, 75–85. (2) Lehnberger, C.; Scheller, D.; Wolff, T. Heterocycles 1997, 45, 2033–2039. (3) De, T. K.; Maitra, A. Adv. Colloid Interface Sci. 1995, 59, 95–193. :: (4) Evans, D. F.; Wennerstom, H. The colloidal domain, 2nd ed.; VCH Publishers: New York, 1999. (5) Chen, S.-H.; Chang, S.-L.; Strey, R. J. Chem. Phys. 1990, 93, 1907–1918.
Langmuir 2009, 25(14), 7927–7931
Published on Web 05/12/2009
DOI: 10.1021/la900592x
7927
Article
Bufe and Wolff
Figure 1. Pseudobinary phase diagram (temperature T versus n-decane fraction by weight) of an AOT-n-decane-deuteriumoxide microemulsion at a constant AOT fraction of 12% by weight (modified from ref 5). LR, lamellar lyotropic liquid crystalline phase; 2φ, upper 2-phase region; 2φ0 , lower 2-phase-region. The illustration of the monophasic region represents areas of o/w, bicontinuous, and w/o structures. At the percolation line, the transition from the w/o to the bicontinuous structure takes place. Scheme 1. AOT
of N-methyl-2-quinolone in AOT microemulsions,8c which takes place upon reirradiation at wavelengths around 250 nm. At this wavelength both the monomer and the dimer absorb light so that the reversion cannot become complete. A photostationary equilibrium of monomers and dimers is formed, depending on absorption coefficients and quantum yields. Accordingly, the percolation point is not shifted back to its value in the presence of pure monomeric N-methyl-2-quinolone. Nevertheless, a partial reversion of the percolation point shift was observed (see Figure 5 in ref 8c). The effect, however, could not be cycled many times due to photochemical side reactions of N-methyl-2-quinolone, of its dimers, or of AOT, which all absorb 250 nm light. Similar behavior can be expected when the photodimerization of anthracenes8b or acridizinium cations8a is reverted by short wavelength light. One example for (ii) was reported so far,9 which is however way off practical use for a switch: the trans-cis photoisomerization of methylhydroxystilbazolium bromide (HSB);inducing a quite small shift in Tp;was thermally inverted by heating the samples to 60 °C for 40 h, cf. Figure 2, and the conductivity versus temperature curve induced by the solubilizate in the trans-form was regained. In these previous examples, the (mostly incomplete) reversion of the percolation point shift either requires reirradiation at another wavelength or time-consuming heating and recooling of samples, both quite long-winded for a switch. In order to exploit the effects for practical conductivity switching (without using a mechanical switch), it was thus desirable to find a 7928 DOI: 10.1021/la900592x
Figure 2. Conductivity vs temperature diagrams (modified from ref 9) for a microemulsion of the composition AOT/isooctane/ water = 1:3:2 (weight ratios) without solubilizate (0, Tp = 27 °C, shoulder and maximum above 35 °C are due to entering the upper two-phase region,8c,9 see Figure 1), in the presence of trans-methylhydroxystilbazolium bromide (HSB, R = methyl) at 0.5% (weight methylhydroxystilbazolium bromide/weight microemulsion = 0.005) (O, Tp = 36 °C), after photoisomerization (3, Tp = 40 °C), and after thermal (40 h at 60 °C) reisomerization (+, Tp = 36 °C).
percolation point shifting photoreaction that is thermally reversible at room temperature within reasonable times. We will show below that azobenzene as a solubilizate in AOT-isooctanewater microemulsions (behaving similar to the AOT system in Figure 1) is suitable for this purpose. Azobenzene was chosen because it was expected to meet the demands: it isomerizes photochemically, and it reisomerizes thermally and faster10-12 than HSB. Moreover, it is commercially available. Azobenzene undergoes photochemical transformations from the trans-form to the cis-form and vice versa (Scheme 2), depending on the irradiation wavelength. As the trans-form is thermodynamically more stable, the cis-form can also react to the trans-form thermally in the dark.10 As will be shown below, trans- and cis-azobenzene do induce different percolation temperatures, and thus, the shift in Tp induced by the photochemical transformation from the trans-form to the cis-form is reverted during the thermal back reaction from cis to trans. Azobenzene was reported to survive multiple photochemical isomerizations and thermal reisomerizations without fading, for example, in the photo-orientation of liquid crystalline polymers.11 In strongly acidic aqueous solutions, however, a very slow side reaction was found.12
2. Experimental Section 2.1. Chemicals and Solutions. Azobenzene was bought from Merck-Schuchardt (specified purity > 98%); sodium bis-2-ethylhexylsulfosuccinate (aerosol OT, AOT; purchased from Fluka, g99.0%) and isooctane (purchased from AppliChem, g99.5%) were used as supplied. Microemulsions were prepared by mixing AOT, isooctane, and doubly distilled water in the desired mass ratio, cf. ref 8b. 2.2. UV-Vis Spectroscopy. UV-vis absorption spectra were taken on an IKS Optoelektronik/Polytec model X-dap spectrophotometer, if necessary after diluting the samples with an appropriate amount of pure microemulsion. :: (10) Rau, H. In Photochromism - Molecules and Systems; Durr, H., BouasLaurent, H., Eds.; Elsevier: Amsterdam, 1990; Chapter 4. :: (11) (a) Stumpe, J.; Lasker, L.; Fischer, T.; Rutloh, M.; Kostromin, S.; :: Ruhmann, R. Thin Solid Films 1996, 284-285, 252–256. (b) Fischer, T.; Lasker, :: L.; Czapla, S.; Rubner, J.; Stumpe, J. Mol. Cryst. Liq. Cryst. 1997, 298, 213–220. (c) Fischer, T.; Menzel, H.; Stumpe, J. Supramol. Sci. 1997, 4, 543–547. (12) Griffiths, J. Chem. Soc. Rev. 1972, 1, 481–493.
Langmuir 2009, 25(14), 7927–7931
Bufe and Wolff
Article
Scheme 2. Photoisomerization of Azobenzene
2.3. Conductivity Measurement, Tp Determination, and Irradiation. Conductivities were measured on a Mettler Toledo SevenMulti conductivity meter operated at 50 Hz using an inLab 710 measuring cell. As described previously in detail,8b the onset of conductivity in conductivity versus temperature curves was taken as percolation temperature Tp. (Percolation mechanisms6,7,13 are not discussed here.) In order to determine Tp reproducibly, a threshold value was set: passing 100 μS cm-1 in conductivity upon heating indicates percolation (several authors use mathematical procedures for defining the percolation temperature from κ versus T plots6,7,13 that may yield Tp values differing by a few degrees as discussed previously8b). It should be mentioned that different charges of AOT may exhibit individual percolation lines, even when purchased from the same supplier. This is probably due to the sensitivity of AOT toward hydrolysis3 and to the fact that AOT as supplied is not a stereochemically pure substance (three asymmetric C-atoms cause four possible diastereomeric structures that can show differing physical properties). At the composition investigated (AOT/isooctane/water = 1:2.5:2 by weight), Tp of the microemulsion is (30 ( 3) °C. Percolation temperatures were determined with an experimental standard deviation of the mean: |ΔTp| e 0.5 K (from four to six measurements). For determining Tp in nonirradiated microemulsions (containing azobenzene), the samples were protected from daylight in order to prevent photoisomerization. Irradiations of ca. 20 cm3 samples were performed using the apparatus sketched in Figure 3, which simultaneously allowed irradiation and conductivity measurements. A 200 W high pressure xenon-mercury lamp (AMKO) served as the irradiation source. Solutions of solubilizates in microemulsions were deaerated by bubbling with argon and irradiated at the desired temperature through the gas/liquid interface employing light filtered by water and Duran glass (λ > 310 nm) for photoisomerization. The experiment which resulted in the series of absorption spectra shown in Figure 5 was performed in an apparatus differing from that in Figure 3: A cuvette (d = 0.200 cm) containing an AOT-isooctane-water microemulsion (1:2.5:2 w/w/w) with solubilized azobenzene (0.0067%; mass ratio mAB/mME = 0.000067, c = 0.3 mmol dm-3) was irradiated with monochromator (AMKO) filtered light at λ = (316 ( 5) nm for 10 min. Subsequent to the irradiation, absorption spectra were taken after various time intervals (see Figure 5) for observing the thermal back reaction.
3. Results 3.1. Influence of Azobenzene on Tp. Throughout this work, we used microemulsions of the composition AOT/isooctane/ water = 1:2.5:2 by weight. At this composition, the microemulsions (pure and with azobenzene) exhibit Tp within an easy-tohandle temperature range around 30 °C. Figure 4 reveals that adding azobenzene to this microemulsion leads to a linear increase of Tp, which was detected up to an azobenzene content of 2% by weight. At a content between 2 and 5% by weight, the curve deviates from linearity. (13) (a) Eicke, H. F.; Thomas, H. Langmuir 1999, 15, 400–40. (b) Paul, S.; Bisal, S.; Moulik, S. P. J. Phys. Chem. 1992, 96, 896–901.
Langmuir 2009, 25(14), 7927–7931
Figure 3. Apparatus for irradiating the samples and simultaneous measuring of electrical conductivity. (1) Lamp housing with high pressure Hg-Xe lamp, (2) heat filter (beaker filled with water), (3) cutoff filter (glass), (4) sample cell made of glass with two openings (one for the light beam and the other for the measuring cell), (5) sample (microemulsion with or without solubilized azobenzene), (6) magnetic stirrers, (7) measuring cell for recording electrical conductivity and temperature, (8) thermostatted water bath, (9) connection to conductivity meter, and (hν) light beam from the lamp to the sample.
Figure 4. Dependence of percolation points Tp on the concentration of azobenzene in a microemulsion of the composition AOT/ isooctane/water = 1:2.5:2 (weight ratios). The difference between the percolation temperature of the microemulsion with solubilized trans-azobenzene Tp (ME/AB) and that of the pure microemulsion Tp (ME) is taken as ordinate.
3.2. Influence of trans-cis Photoisomerization on Tp. Although the thermal reversibility of the azobenzene photoisomerization has been proven in a variety of solvents,12 we tested AOT microemulsions as a solvent in this respect via UV-vis absorption spectroscopy. In Figure 5, spectra taken before and after irradiation (in a 0.2 cm cuvette) as well as during the thermal back reaction are shown. Under these conditions (room temperature, azobenzene concentration c = 0.3 mmol dm-3), the back reaction takes about 7 days for completion, which is in fair agreement with literature data obtained for azobenzene in chloroform,14 that is, in a polar environment to be compared with the polar interfacial region of microemulsions where the thermal back reaction of the highly polar cis-azobenzene can be expected to take place. (14) Talaty, E. R.; Fargo, J. C. Chem. Commun. (London) 1967, 65–66.
DOI: 10.1021/la900592x
7929
Article
Bufe and Wolff
Figure 5. UV-vis absorption spectra of azobenzene in a w/o microemulsion (AOT/isooctane/water = 1:2.5:2 by weight) before and at various times after irradiation, that is, at various degrees of thermal reisomerization, taken at room temperature; concentration of azobenzene c = 0.3 mmol dm-3, mass ratio mAB/mME = 0.000067.
Figure 6. Conductivity versus temperature diagrams for a microemulsion of the composition AOT/isooctane/water = 1:2.5:2 (weight ratios) in the presence of azobenzene at 0.2% (weight azobenzene/weight microemulsion = 0.002), taken before irradiation, that is, in the dark (O), under irradiation (3), and in the dark 6 days after irradiation (+). The difference between the actual temperature T and the percolation temperature of the microemulsion with solubilized azobenzene in the dark Tp (ME/AB/dark) is taken as abscissa.
In Figure 6, conductivity versus temperature plots in the presence of 0.2% azobenzene in the dark and under irradiation are shown, clearly revealing that the solution under irradiation (cis-rich) exhibits a percolation temperature decreased by ca. 6 K. After switching off the light, the conductivity versus temperature curve reverts to the one measured in the dark. Thus, at a fixed temperature (between the two percolation temperatures), the conductivity can be switched on and off by switching the irradiation light on and off as indicated by the arrows in Figure 6. At azobenzene concentrations other than 0.2%, that is, at 2% and at 0.05%, the difference of percolation temperatures in the dark and under irradiation was smaller. 3.3. Reversible Switching of Conductivity. Multiple on/off switching of conductivity is demonstrated in Figure 7 for samples in which two, five, and ten on/off cycles were run. Although considerable scatter is evident in the maximum regions, a fading of the effect cannot be observed. The time needed for switching on and off conductivity (evaluated from Figure 7) was about 6 min always, as listed in Table 1. Some heating of the samples upon switching on the light is unavoidable in our apparatus due to heat emission by the lamp in 7930 DOI: 10.1021/la900592x
Figure 7. Conductivity versus time diagrams for three individual microemulsions of identical composition AOT/isooctane/water = 1:2.5:2 by weight in the presence of 0.2% azobenzene (azobenzene/ microemulsion = 0.002 w/w) during repeated on/off switching of light for irradiation. Rising conductivity follows switching on the light, falling conductivity follows switching off the light. (a) Two switching cycles; (b) five switching cycles; (c) ten switching cycles. Table 1. Times Needed for Switching on and off Electrical Conductivity during the Switching Cycles Represented in Figure 7 experiment according to Figure 7
a
b
c
a, b, c
number of switching cycles time for switching on conductivity mean value/min experimental standard deviation/min experimental standard deviation of the mean/min
2
5
10
17
4.8 0.2 0.09
6.6 1.5 0.7
6.3 1.6 0.5
6.2 1.5 0.4
6.1 1.6 1.1
9.7 4.7 2.1
4.9 0.7 0.3
6.4 3.3 0.8
time for switching off conductivity mean value/min experimental standard deviation/min experimental standard deviation of the mean/min
addition to the emission of UV and visible light. This does, however, not lead to temperatures exceeding the percolation point in the dark as demonstrated in Figure 8. Langmuir 2009, 25(14), 7927–7931
Bufe and Wolff
Article
Figure 9. Assumed location of azobenzene molecules (b) in the interface region of an AOT w/o microemulsion (O, polar head groups of AOT molecules). (a) Oil-water interface in a pure microemulsion; (b) microemulsion with solubilized trans-azobenzene; (c) microemulsion with solubilized cis-azobenzene.
Figure 8. Conductivity (0) and temperature (b) versus time diagram for a microemulsion of the composition AOT/isooctane/ water = 1:2.5:2 (weight ratios) in the presence of 0.2% azobenzene (weight azobenzene/weight microemulsion = 0.002) during two on/off switching cycles of light for irradiation. The horizontal line at 30.2 °C represents Tp in the dark.
4. Discussion 4.1. Origin of Tp Shifts. We can expect a preferred location of azobenzene in the interface region of microemulsions because of the strong effects on percolation temperatures. As AOT molecules exhibit a conical shape (see formula in Scheme 1, and consider the packing15 parameter > 1), they easily arrange around water droplets and give rise to a quite extended w/o region in phase diagrams (Figure 1). Considering the much higher polarity of cis-azobenzene, we can expect this isomer residing more near to the ionic head groups of AOT than the transazobenzene. This consideration is pictured in Figure 9. It can be seen that molecules in the trans-form increase the hydrophobic volume of AOT and thus the curvature of the oil-water interface (Figure 9b) while cis-molecules decrease the curvature (Figure 9c). The latter effect can facilitate the transition from the w/o state to the percolated (spongelike) state, as a flattening of the water-oil interface is required. In keeping with this interpretation is the observation that the wettability of monolayers increases when azobenzene moieties within the layers are photoswitched to the cis-form (and vice versa).16 4.2. Reversible Switching. When comparing Figures 5 and 7, one realizes that the electrical conductivity switches faster than the thermal cis-trans isomerization proceeds. The reason is that only a part of the photochemically produced cis-azobenzene molecules has to reisomerize in order for the system to depercolate. It follows that switching times can be further optimized (shortened) in that the starting microemulsion is set more near to (15) Israelachvili, J. Intermolecular and surface forces, 2nd ed.; Academic Press, Inc.: San Diego, 1991. (16) Jiang, W.; Wang, G.; He, Y.; Wang, X.; An, Y.; Song, Y.; Jiang, L. Chem. Commun. 2005, 3550–3552.
Langmuir 2009, 25(14), 7927–7931
the percolation point under light by either varying concentrations or temperature. The conductivity switching amplitude, however, may be smaller then. The fact that a switching amplitude maximum was observed at 0.2% azobenzene concentration can be rationalized as follows: at small concentrations, the initial trans-azobenzene effect is smaller (Figure 4); at high concentrations, the limited volume of the AOT layer becomes overcrowded, so that a substantial part of the solubilizate is dissolved either in the bulk organic phase (transform) or in the water droplets (cis-form) and cannot fully act with respect to the specific effects on Tp. This may also explain the deviation from linearity in Figure 4 at higher azobenzene concentrations. The scatter observed in the maximum regions of Figure 7 is due to fluctuations of domains in the bicontinuous phase and also to stirring problems in the apparatus used. Especially around the conductivity measuring cell (item 7 in Figure 3), the exchange of material by stirring needs improvement.
5. Conclusions and Outlook Reversible on/off switching of electrical conductivity in AOT microemulsions is possible without changing temperature or composition. The reversible transition from the w/o to the percolated solution structure is induced by the trans-cis photoisomerization of solubilized azobenzene (acting as a large response trigger) and reverted during thermal reisomerization. This is brought about simply by exposing samples to light that is switched on and off, that is, without turning a mechanical contact in the conductivity switching device. For possible applications, it is advantageous that the microemulsion is composed of commercially available chemicals, that is, the synthesis of special surfactants or solubilizates is not necessary. Future efforts will concentrate on finding other and faster thermally reversible photoreactions that are capable of shifting percolation points and on finding further percolating microemulsion systems to substitute the expensive AOT. Acknowledgment. Financial support by the Deutsche Forschungsgemeinschaft (DFG) is gratefully acknowledged. :: We thank Mrs. A. Gopfert and Mrs. U. Georgi for technical assistance.
DOI: 10.1021/la900592x
7931
