J. Chem. SOC., Faraday Trans. I, 1988, 84(7), 2241-2245 Localized Excess Electrons in Solubilized Water Clusters in Aerosol OT-n-Heptane Solutions Mahmoud H. Abdel-Kader*t Faculty of Applied Science, Arabian Gulf University, P.O. Box 26671, Manama, Bahrain Peter Krebs Institut f u r Physikalische Chemie und Elektrochemie der Universitat Karlsruhe, Federal Republic of Germany Excess electrons have been produced by flash photoionization of OH- in the water pool of reversed micelles. The optical absorption spectra of these electrons have been investigated as a function of the water cluster size. The results indicate that the shape of the absorption spectra as well as lifetimes of the excess electrons are in partial contradiction of the results reported in the literature. In some recent inve~tigationsl-~ excess electrons were used as a probe to obtain structural information on surfactant [sodium bis(2-ethylhexylsulphosuccinate), aerosol OT, AOTI-solubilized water pools in non-polar solvents (so-called reversed micelles).The number of water molecules in such a pool depends mainly on the molar ratio w = [H,O]/[AOT], and can easily be varied in a wide range. This will result in reversed rriicelles of different sizes having an almost monodispersive size di~tribution.~. Localized electrons2 in the water pools of the reversed micelles can be produced radiolytically or photolytically. The properties of these excess electrons depend strongly on the water content of the reversed micelles. The most salient results of these studies, some of which are controversial, are listed below.(i) With decreasing water content the yield of pulse radiolytically produced excess electrons is reduced." ' 9 ' Below a critical size of the water core (w < 5) localized electrons cannot be observed.'? (ii) For w > 15 the spectrum of the electrons in the water pool was found to be identical with that of hydrated electrons in the bulk water phase.* By lowering w, the maximum of the optical absorption spectrum of the excess electrons in the micellar sollutions becomes hypsochromically shifted compared with the spectrum of hydrated electrons in bulk water.l, ' v 4 Fendler and co-worker~,~ however, did not find any shift, even for very small water pools (w z 5). (iii) The bandwidth of the broad absorption spectrum increases with decreasing size of the water core,4 whereas the results of Wong et a1.l and Fendler and co-workers3 show a dramatic decrease of the bandwidth.The latter result was taken to indicate an ice-like structure of the water molecules in the (iv) The polarity of relatively large water clusters of radius R = 73 A (corresponding to w z 49) is still lower than that of bulk water.l It is not expected that this fact should be without any influence on the spectroscopic properties of the excess electrons, even in somewhat smaller water pools (w 2 15).4 7 Permanent address : Department of Chemistry, Faculty of Science, Tanta University, Tanta, Egypt. $ We use the term localized electrons instead of hydrated electrons since, especially in the small water pool of reversed micelles, where the number of Na+ ions is comparable to the number of H,O molecules in the pool, the electron is not hydrated in the literal sense.224 12242 Soluhilized Water Clusters (v) The lifetime of the electrons produced by pulse radiolysis was observed to be of the order of 100-200 ns1T4 and was strongly reduced by decreasing the water content of the reversed mi~elles.~ In contrast, the lifetime of photolytically produced electrons was found to be of the order of microseconds, increasing with decreasing water pool size.:' In both cases interpretation of the contradictory experimental observation was given. Here we report the results on the absorption spectra of excess electrons in the water pool of reversed micelles produced by laser flash photoionization of tetrabutyl- ammonium hydroxide (TBAOH).This compound is most probably solubilized at the micelle interface, thus leading to the formation of OH- in the water pool. The advantage of using OH- as electron donor in the photoionization processes is that it does not produce any additional transients in the wavelength region where the localized electron absorbs light. Experiment a1 Sodium bis(2-ethylhexylsulphosuccinate) (Fluka 98 %) was purified by column chromatography on activated charcoal as described by Calvo-perez et al.,3 using the mixing volume ratio (1 : 5 ) V/V cycholhexane/methanol. AOT was dried under vacuum at 320 K, n-heptane (Merck, for spectroscopy), tetrabutylammonium hydroxide (Fluka, 98 % ; 40 % in H,O) and KI (Merck, extra pure) were used as received.mol dm-3] in inverse micellar solutions of 0.1 mol dm-3 AOT in n-heptane containing different amounts of triply distilled water (the last distillation step was performed in a quartz apparatus) were prepared under vacuum in a special storage vessel with a 1 em spectrosil quartz cell (Hellma GmbH). These solutions were carefully degassed by several freeze-pumpthaw cycles. The excess electrons in the water pool were produced by photoionization of OH-, using the light pulse of a frequency quadrupled Nd-YAG laser (J. K. Lasers, System 2000, II = 265 nm, pluse duration 15 ns, pulse energy 10-25 mJ) according to (OH-)+hv -+ (HO')+e;& The absorption of e;",, was monitored by a pulsed high-pressure xenon arc lamp, detected by a fast Si-photodiode (modified UDT 600, United Technology), and displayed on a Tektronix 7633 storage oscilloscope.The experimental set-up is described el~ewhere.~ All experiments were carried out at room temperature. KI (4 x mol dm-3) in water and TBAOH [(2-8) x Results and Discussion In order to confirm the reproducibility of our experimental set-up we have measured the spectra of hydrated electrons produced by laser flash photolysis of I- and OH- in pure water. The results reveal that A,,, and the width of the spectrum of e& in bulk water was in agreement with the most reliable spectrum of Michael et al.,u where the electrons were generated by pulse radiolysis. Therefore possible variations of the shape of the electron spectra in inverse micellar solutions can be studied with this experimental set-up.The normalized spectra of localized excess of electrons in AOT-H,O-n-heptane solutions produced by photoionization of TBAOH are shown in fig. 1 for different H,O-AOT concentration ratios w in comparison with the results of other groups. Solutions containing 0.1 rnol dm-3 AOT, 2.22 mol H,O (w = 22.2) and 5 x lop3 mol dmP3 TBAOH are compared with the pulse-radiolytic result of Wong et a1.l for w = 49. Both spectra show approximately the same position of the maxi- mum as the hydrated electron in pure water (hv,,, = 1.72 eV, &,, = 720 nm), but their widths are strongly reduced. This finding is, however, in contradiction to the observations of Pileni et al.,4 who did not observe any decisive deviation from the shape of the hydrated electron spectrum in pure water.M .H. Abdel-Kader and P . Krebs 2243 1.0 - - - W G - - P, 2 - 2 0.5- 2 - - - D .* Y - m - - - - - ' # ' - I I I I I 400 500 600 700 800 900 1000 wavelength/nm Fig. 1. Normalized absorption spectra of excess electrons in large water pools of inverse micellar solutions in comparison with that of the hydrated electron in the bulk water phase (-.-). (-O-) pulse radiolysis: 6% v/v H,O in 3% w/v AOT-heptane solutions (MI = 49).l (-O-) This work: photoionization of 5 x mol dm-3 TBAOH in a solution of 0.1 mol dm-3 AOT and 2.22 mol dm-3 H,O in n-heptane ( w = 22.2). The different points at each wavelength demonstrate the reproducibility of these measurements. The insert shows the decay of the optical absorption with time at R = 700 nm (the small peak at t = 0 is due to scattered laser light).0.0 I ' 1 I I I I 1 1.0 0.0 t t t *$ + * t- * -I 1 1 V I I 1 i I 400 500 600 700 800 900 1000 wavelength/nm Fig. 2. Normalized spectra of electrons in small water cores of reversed micelles. (-F) Pulse radiolysis: 1 Yo v/v H,O in 3 YO w/v AOT-heptane solutions ( w = S).' (. + - ) photoionization of 1 x rnol dm-3 phenothiazine in a solution of 0.25 mol dm-3 H,O and 0.05 mol dm-3 AOT in heptane ( w = 5).3 (-.-) Pulse radiolysis: AOT-H,O-iso-octane solution with w = 5.4 (-O-) This work : photoionization of 5 x lop3 rnol dm-3 TBAOH in a solution of 0.55 mol dm-3 H,O and 0.1 mol dmP3 AOT in n-heptane ( w = 5.5). Fig. 2 compares our results for w = 5 with the results for w = 5-8 from several other groups (some of which used iso-octane instead of n-heptane).The following observations can be made. (1) All spectra shown, except these of Fendler and co-worker~,~ indicate a shift of hmaX to higher energies. One has to take into account that the spectrum published by Fendler and co-workers was observed by laser flash photolysis of phenothiazine in2244 Solubilized Water Clusters inverse micellar solutions. Here the wavelength region of the ' hydrated ' electron spectrum overlaps with the absorption of the phenothiazine cation radical. Subsequently, it was not easy to determine the electron spectrum exactly from these measurements. (ii) Wong et al. determined a shifted spectrum, whose shape has completely lost the typical characteristics of the excess electrons spectrum. (iii) The position of the maximum of the electron spectrum determined by Pileni et al.is almost the same as that of Wong et al. and ours. However, the bandwidth, which is in qualitative agreement with our experimental result, is twice as high as that Wong et al. (iv) It was asserted by different authors1Y4 that the absorption of excess electrons could not be observed in a solution with w < 5. In this case it was assumed that all water molecules participate as part of the solvation shell of the Na+ counterions of the surfactant molecules. In the pulse-radiolytic experiments the disappearance of any absorption was explained by the fact that there are no further water molecules in the pool to trap the electrons originally produced in the non-polar phase of the solution. However, our experiments with w z 1 yielded a broad transient absorption in the wavelength region 500 d L/nm < 700 (lifetime = 2ps).This may be ascribed to the absorption of excess electrons in an "a+-H,O matrix'. This interpretation is also supported by the experimental fact that the observed transient absorption disappears in the presence of typical electron scavengers such as N,O and 0,. In contrast to the results of Pileni et u Z . , ~ our measurements demonstrate that even for w = 22.2 the absorption spectrum of excess electrons in the water pool of reversed micelles is different from that in the bulk water (fig. I). This spectroscopic result is also supported by several independent experiments. Wong et al.' found from their n.m.r. studies that the micellar water is highly structured at least for w 5 8.In this case all water molecules are strongly bonded to the Na+ counterions of AOT. However Bakale e? al." showed (in an AOT-H,O-iso-octane solution) that the attachment of electrons (originally produced by pulse radiolysis in iso-octane) to the water pools does not become diffusion-controlled until w > 30. This implies that the properties of the excess electrons in the water pool should be different from those in the bulk water phase for w < 30. For w = 20 the mean Na+ concentration in the water pool is 2.24 mol dm-3. According to Kreitus et a1.l' only a small shift of the electron absorption maximum to higher energies is to be expected (e.g. for electrons in an aqueous 2 mol dmP3 LiCl solution Ahvmsx 5 0.1 eV with respect to hv,,, = 1.72 eV for the hydrated electrons).Indeed, apart from a decrease in the bandwidth, we have observed no noticeable shift of the spectrum. Therefore, it is concluded that the distribution of the Na+ ions is inhomogeneous and that the hydrated Na+ ions are located near the micelle interface. The rest of the water pool, however, seems to be different from bulk water, thus influencing the width of the broad electron absorption band. As shown in fig. 1, the observed narrowing of the absorption spectra of e;",, can be ascribed to the changes in the properties and organization of the water molecules in the pool which possesses a more ordered environment compared to the bulk water. Such narrowing is also rationalized in terms of increasing the viscosity with decreasing w value^.^ However, with a small water pool (w < 5 ) the bandwidth is broader than :hat observed for w = 22.2, in accordance with the experimental findings of Pileni et ~ 1 .~ This broadening may be attributed to the absorption of excess electrons in an Na+-H,O matrix' exhibiting a broad spectrum, but still narrower than that measured in bulk water. For w > 15 Pileni et al. determined a w-independent lifetime of ca. 230 ns for the localized electrons (they disappear according to a pseudo-first-order reaction). Hence the half-life is at least two orders of magnitude lower than in bulk water.4 Pileni et al. assumed that the electrons were reacting with AOT, because by diminishing w (for w < 15), i.e. by decreasing the distance between the electron and the surfactantM.H. Abdel- Kader and P. Krebs 2245 molecules, the electron lifetime is strongly reduced. However, Fendler and co- workers3 reported that the lifetime increases with decreasing size of the water pool of the inverse micelle. Moreover, the decay time of the excess electrons was found by them to be of the order of microseconds. The latter result is also supported by our own measurements (fig. 1). Even at very low ratios ( w = 1) the lifetime (measured at ;1 = 700 nm) is of the order of 2,us or greater. This would tend to contradict the interpretation of the fast decay of the electrons found by Pileni et aL4 The observed difference in the elo,, lifetime may be attributable to impurities which are reactive with electrons.Recently Pileni and co-workers12 reported the lifetime of e& for w 3 15 as 500 ns. Thus it can be concluded that the elo,, lifetime is very sensitive to the solution conditions and also to the method of production. It is worthwhile mentioning that our results, obtained by photoionization of OH- in aqueous solutions are similar to those reported using the pulse-radiolysis technique.* However, the results in microheterogeneous media seem to be very sensitive to the elo,, production method as well as the solution conditions in which the pH value is notoriously difficult to control. We thank Dr A. M. Braun, EPFL, Lausanne, for pointing out to us that TBAOH is a suitable ‘electron donor’ in the photoionization experiment, and Dr V. Giraud for assistance with the experimental work. Financial support by the Internationales Seminar an der Universitat Karlsruhe and by the Deutsche Forschungsgemeinschaft is acknowledged. References 1 M. Wong, M. Gratzel and J. K. Thomas, Chem. Phys. Lett., 1975, 30, 329. 2 J. K. Thomas, F. Grieser and M. Wong, Ber. Bunsenges. Phys. Chem., 1978, 82, 937. 3 V. Calvo-Perez, G. S. Beddard and J. H. Fendler, J. Phys. Chem., 1981, 85, 2316. 4 M. P. Pileni, B. Hickel, C. Ferradini and J. Puncheault, Chem. fhys. Lett., 1982, 92, 308. 5 M. Zulauf and H. F. Eicke, J. fhys. Chem., 1979, 83, 450. 6 P. D. I. Fletcher and B. H. Robinson, Ber. Bunsenges. fhys. Chem., 1981, 85, 863. 7 St. Jaenicke and P. Krebs, J. Phys. Chem., 1980, 84, 1119. 8 B. D. Michael, E. J. Hart and K. H. Schmidt, J . Phys. Chem., 1971,75, 2798. 9 M. Wong, J. K. Thomas and T. Nowak, J. Am. Chem. SOC., 1977, 99, 4730. 10 G . Bakale, G. Beck and J. K. Thomas, J. Phys. Chem., 1981, 85, 1062. 11 I. V. Kreitus, V. A. Benderskii, A. G . Krevenko and Yu. E. Tiliks, J. Electroanal. Chem., 1982, 133, 345. 12 C. Petit, P. Brochette and M. P. Pileni, J. Phys. Chem., 1986, 90, 6517. Paper 71741 ; Received 23rd April, 1987 74 FAR 1