Molecular Diffusion and Fluorescence Energy-Transfer Studies in Thin

Mar 27, 1995 - Thin surfactant films have been formed by dip-coating glass slides in a solution of .... (13) Papoutsi, D.; Líanos, P. Langmuir 1995, ...
1 downloads 0 Views 765KB Size
Langmuir 1995,11, 4355-4360

4355

Molecular Diffusion and Fluorescence Energy-Transfer Studies in Thin Surfactant Films Despina Papoutsi, Vlasoula Bekiari, Elias Stathatos, and Panagiotis Lianos" Engineering Science Department, University of Patras, GR-26500 Patras, Greece

Andre Laschewsky Departement de Chimie, Universite Catholique de Louvain, B-1348 Louvain-la Neuve, Belgium Received March 27, 1995. I n Final Form: September 6, 1 9 9 P Thin surfactant films have been formed by dip-coating glass slides in a solution of reversed micelles containing titanium isopropoxide. The alkoxide is slowly hydrolyzed in the presence of reversed micelles since hydrolysis competes with hydration of surfactant polar groups. Adhesion of the surfactant on the glass slide is assisted by incompletelyhydrolyzed alkoxide through the following possible mechanism: the alkoxide adheres by -Si-0-Ti- bonds and the surfactant follows by hydrophobicattraction to the isopropyl groups. The films are transparent and visually uniform. Three surfactants, forming reversed micelles, have been investigated: one nonionic, Triton X-100; one anionic, AOT; and one cationic, hexadecyldimethylbenzylammonium chloride. The environment provided by the surfactant film has been studied by fluorescenceprobing. In particular, we have analyzed pyrene excimer formation as well as energy transfer between pyrene, acting as donor, and coumarin-153 or N-n-heptyl-4-(((dimethylamino)phenyl)ethenyl)pyridinium bromide, acting as acceptor. In Triton and hexadecyldimethylbenzylammonium chloridefilms, pyrene excimer formation is diffusion-controlledwhile in AOT films excimer largely comes from pyrene aggregation. Pyrene excimer formation capacity decreases in the presence of cosulubilized poly(viny1 methyl ether) chains. Energy transfer data indicate that coumarin-153 is randomly distributed in the films,but N-n-heptyl-4-(((dimethylamino)phenyl)ethenyl)pyridiniumbromide is not randomly distributed but shows a tendency to aggregate. Generally speaking, the benzyl-group-bearingsurfactants form fluid structures while AOT may provide a restricted low dimensional environment.

1. Introduction The study of organic functional groups self-assembled on solid substrates has attracted increasing attention during the last decade because of a multitude of possible applications. Most attention has been paid to highly ordered monolayers, owing to their non-linear optical properties,1,2 to their function in the construction of microelectrode^,^ and to several other high-technology application^.^ Stable, structured surfactant films, however, also have similar applications, even when they do not possess the high order that qualifies monolayer^.^ Thus biologists are intrigued by the resemblance of lipid membranes to thin organic films since they provide a basis for biomimicry and for understanding several biological functions. In this work, we have prepared and studied thin films composed of common surfactants, usually employed for making micelles and microemulsions. Our goal was to investigate their degree of organization in the film as well as the possibility of molecular diffusion and energy transfer there, i.e. the two most basic functions in a biological membrane. Organized molecular assemblies in solution are recognized as successful supports of active dopants, as well as efficient microreactors for the synthesis of novel

* Author to whom correspondence should be addressed: Tel30FAX 30-61-997803; e-mail LIANOS@ UOPVAX.UPATRAS.GR. Abstract published in Advance A C S Abstracts, November 1,

organic or inorganic material^.^-^ It is very intriguing to investigate similar functions in thin films, since the synthesized substances can be then useful as functional composites of the complex assembly. Surfactant films have already been used as host for proteins,1° and this possibility adds new dimensions to the importance of these systems. Most of energy or electron transfer studies in thin surfactant films are carried out in monolayers and Langmuir-Blodgett films.ll We have prepared our surfactant films by a novel procedure, assisted by formation of titanium dioxide through hydrolysis of titanium isopropoxide. Incompletely solubilized alkoxide is anchored on glass slides by forming -Si-0-Tibonds and attracting surfactant onto the slide by the help of alkyl groups.12J3 Titanium dioxide formation is absolutely necessary for obtaining transparent and uniform films, especially, in the case of nonionic surfactants. The procedure is straightforward, easy, and flexible and permits a high loading of the films with active dopants, which can be introduced in the precursor solutions, as explained below. The final product is a porous titania membrane coated with surfactant. This complex material possesses new properties and promises for new applications. Thus of particular interest could be the photosensitization of titania by electron transfer from functionalized surfactants.14

61-997587; @

1995.

(1)Lupo, D.; Prass, W.; Scheunemann,U.; Laschewsky, A.; Ringsdorf, H.; Ledoux, I. J. Opt. SOC.Am. B 1988,5, 300. (2) Yamada, S.; Kawazu, M.; Matsuo, T. J. Phys. Chem. 1994, 98, 3573.

(3) Abbott, N. L.; Robinson, D. R.; Whitesides, G. M. Lungmuir 1994, 10, 2672. (4) Mao, G.;Tsao,Y.-H.;Tirell, M.; Davis, H-T.; Hessel, V.;VanEsch, .T ' Ringsdorf, H. Langmuir 1994, 10, 4174. 5) Suga, K.; Bradley, M.; Rusling, J. F. Langmuir 1993, 9, 3063.

(6) Papadimitriou, V.; Xenakis, A,; Lianos, P. Langmuir 1993, 9, $12. (7) Leong, Y. S.; Candau, F. J. Phys. Chem. 1982, 86, 2269. (8) Lianos, P.; Thomas, J. K. Chem. Phys. Lett. 1986, 125, 299. (9)Petit, C.; Lixon, P.; Pileni, M. P. J. Phys. Chem. 1993,97,12974. (10) Hamachi, I.; Shunsaku, N.; Kunitake, T. J. Am. Chem. SOC. 1990,112, 6744. (11)Budach, W.; Ahuja, R. C.; Mobius, D. Thin Solid Films 1994, 243, 647. (12) Papoutsi, D.; Lianos, P.; Yianoulis, P.; Koutsoukos, P. Langmuir 1994, 10, 1684. (13) Papoutsi, D.; Lianos, P. Langmuir 1995, 11, 1.

0743-746319512411-4355$09.00/0 0 1995 American Chemical Society

4356 Langmuir, Vol. 11, No. 11, 1995

Papoutsi et al.

PYRENE

CH3.CHz 0 CH3- (CH2)3-&H- C H 2 . 0 - C FH-S03' Na+ FH2 CH3- (CH2)3-yH-CH2-O-C,, CH3.CH2 0

AOT

Figure 1. Surfactants and fluorophores used.

2. Materials and Methods Pyrene (Fluka), toluene (spectroscopic grade, Merck), cyclohexane (spectroscopic grade), Triton X-100, titanium isopropoxide, and coumarin-153 (Aldrich), hexadecyldimethylbenzylammonium chloride (HDBAC, BDH), and bis(2-ethylhexy1)sulfosuccinate sodium salt (AOT, Fluka) were used as received. Poly(viny1 methyl ether) (Aldrich) was precipitated from water a t 40 "C. N-n-Heptyl-4-(((dimethylamino)phenyl)ethenyllpyridinium bromide (H7HC) was synthesized by condensation of the corresponding N-alkylated 4-methylpyridines with 4-(dimethy1amino)benzaldehyde and it was purified by repeated recrystallization from ethanol.15 Millipore water was used in all experiments. Mixtures of cyclohexane, Triton, and water were made a t ambient temperature by first dissolving surfactant in cyclohexane and then adding water. The mixture is not clear below 25 "C but it clears out upon addition of titanium isopropoxide under vigorous stirring. Several ratios of the above components have been tried, but optimum conditions were obtained when the final concentrations ofTriton, water and alkoxide in cyclohexane were 0.1, 0.2, and 0.1 M, r e s p e c t i ~ e l y . ' ~ JAfter ~ a few minutes of stirring the solution, a properly cleaned glass (or quartz) slide was dipped into the solution and withdrawn a t a rate of 0.34 m d s . A thin, uniform, transparent film then formed which was left to dry in air. Gravimetric and spectrophotometric (light interference) measurements gave an estimation offilm thickness. It ranges within 100-200 nm. Films are thicker when the speed of withdrawal of the slide from the solution is higher, or when dipping is done later and not immediately after addition of alkoxide, due to the relatively high viscosity of the solution. As a matter of fact, the solution gels within 2 h after alkoxide addition. We have also observed that gelation becomes faster when Triton in its reduced form is used. Gelation rates are discussed in our previous publication12 and they are related to competition between surfactant and alkoxide for water, in the first case for hydration of surfactant hydrophilic group and in the second case for hydrolysis of alkoxy groups. AOT films were made by the same procedure. The optimum concentrations of surfactant, water, and alkoxide in cyclohexane were 0.2, 0.4, and 0.2 M, respectively. In the case of HDBAC, which is not soluble in cyclohexane, we used toluene as the solvent while the respective surfactant, water, and alkoxide concentrations were 0.05, 0.1, and 0.15 M. Pyrene, H7HC, and coumarin were introduced by solubilization in cyclohexane before alkoxide addition. H7HC was solubilized only in the presence of the oppositely charged surfactant AOT. In all other cases it was insoluble at the concentrations used in this work. All dyes were carried along with the rest of the material forming the film, during withdrawal ofthe slide from the solution. (14)Nazeeruddin, M. K.; Kay, A.; Rodicio, I.; Humphry-Baker, R.; Muller, E.; Liska, P.; Vlachopoulos, N.; Gratzel, M. J.Am. Chem. SOC. 1993.115. 6382. (15) Lunkenheimer, K.;

1992,89,239.

Laschewsky, A. Prog. Colloid Polym. Sci.

a

u.vu

300

LOO

nm

500

Figure 2. Absorption spectra (reference air): (a) (-) 0.05 M HDBAC in toluene, (- - -1 0.2 M Triton X-100 in cyclohexane;

(b) (-1 HDBAC film, (- - -1 Triton film without pyrene, (. . .) Triton film with pyrene (4mM in the original solution). Glass (or quartz) slides were cleaned by using three different methods, i.e. sulfochromic solution, saturated KOH solution, or "piranha solution".l6 All three methods were equally good in our case, so we used only a sulfochromic solution which was more convenient to us. Absorption measurements were made with a Perkin-Elmer Lambda 15spectrophotometer, fluorescence measurements with a home-assembled spectrofluorometer using Oriel parts, and lifetime measurements with the photon-counting technique using a home-made hydrogen flash and ORTEC electronics. Interference filters (Oriel)were used for excitation (335nm) and monomer pyrene emission (380 nm) while a glass filter (Oriel) was used for excimer emission.

3. Results and Discussion

3.1. Spectrophotometric Characteristics of the Materials Forming the Films. The chemical s t r u c t u r e of the s u r f a c t a n t s and the p r o b e s used in the present w o r k are s h o w n in Figure 1. Figure 2 a s h o w s the a b s o r p t i o n s p e c t r u m of hexadecyldimethylbenzylammonium chloride

(HDBAC) and T r i t o n X-100 in the solvents and at the c o n c e n t r a t i o n s used t o prepare p r e c u r s o r solutions ( t h e optical path is 1cm). AOT does n o t a b s o r b in the s p e c t r a l region studied in this w o r k . The corresponding spectra of the films, o b t a i n e d in the t r a n s m i s s i o n m o d e (optical (16)Ulman, A. A n Introduction to Ultrathin Organic Films, from Langmuir-Blodgett to Self-Assembly;Academic Press: San Diego, CA, 1991.

Molecular Diffusion in Surfactant Films

'1

12

A

300

400

,,-..

Langmuir, Vol. 11, No. 11, 1995 4357

a

500 n m

I

400

500 nm

"

"

'

"

'

l

600 b

300

h

550

650

550

600

Figure 3. Absorption spectra (reference air): (a) (-) 5 pM coumarin-153in cyclohexane, (- - -) 3 pM H7HC in ethanol; (b) AOT films containing (-) coumarin-153(5 mM in the original solution) and (- - -1 H7HC (5 mM in the original solution). 350

r:

050

300 320 k m ) 360 Figure 4. Absorption spectrum of pyrene (-1 in a film obtained from a solution containing 16 mM pyrene and (- - -) in pure cyclohexane containing 20 pM pyrene (optical path = 1 cm).

path of about 200 nm), are seen in Figure 2b. The dominant absorption ( '340 nm) mainly belongs to titania. It is characteristic of the absorption spectrum of this semiconductor and it is also observed in the absence of other material. Of course, surfactant absorption is superposed. Pyrene absorption spectrum in a Triton film is also shown in Figure 2b, for comparison. It is possible that substantial screening from the background light absorption and diffusion might influence absorption and emission of photons by pyrene, especially when the fluorophore concentration is smaller than the one used to obtain the above spectrum. This is one reason, for which fluorescence measurements were done in the reflection mode. Coumarin and H7HC are characterized by relatively high extinction coefficients, as seen in Figure 3, showing absorption spectra in solution and in film. The position of their absorption bands justifies their role as efficient resonance energy acceptors of pyrene fluorescence, as will be seen below. 3.2. Pyrene Excimer Formation. Pyrene was incorporated in the film by solubilization in the original solution a t concentrations ranging between 1and 16 mM. Figure 4 shows a n absorption spectrum of 20 pM pyrene in cyclohexane and of pyrene in Triton film. Comparison of the absorbencies indicates that pyrene concentration in the film, if we assume a film thickness of 200 nm and a n equal extinction coefficient, is around 0.2 M, that is much larger than it was in the original solution (i.e., 16 mM). This happens because cyclohexane is evaporated when the material is deposited on the substrate, and thus the remaining material is more concentrated. The volume then approximately decreases by a factor of 10 by solvent evaporation. The thin film absorption is red-shifted. Otherwise, it possesses all the characteristics of cyclo-

410

Mnm)

470

Figure 6. Uncorrected fluorescencespectra ofpyrene: (A) (-)

in film, original concentration 16 mM; (- - -) in film, original concentration 1mM; (. . .)in pure cyclohexane, 0.5 mM; (B) (-) in film, original concentration 1mM; (- - -) in pure methanol, 10 pM; and (. . .) in pure cyclohexane, 10 pM. Table 1. ZJINI, t l , and t z Values for Pyrene in the Films vs Pyrene Concentration in the Original Solution

Triton Films 1 2 4 8 16

1 2 4

8 16

0.13 0.25 0.50

32 (21) 33 (13) 33 (16) 32 (18) 31 (31)

AOT Films 37 (32) 38 (24) 36 (18) 37 (20) 36 (26)

142 (79) 134 (87) 108 (84) 87 (82) 58 (69)

159 (68) 147 (76) 135 (82) 118 (80) 101 (74)

hexane absorption. There is no indication of pyrene ground-state association which would have been demonstrated with broadening ofthe film ab~orption.'~ The red shift of the maxima of film absorption is indicative of multipolar interactions in a polar environment. Figure 5 shows uncorrected pyrene fluorescencespectra. Figure 5a reveals that a substantial quantity of excimer is formed when pyrene concentration in the original solution was 16 mM. No excimer formation was observed with 1mM pyrene. At concentrations between these two extreme values there exists a progressive increase of excimer emission with original pyrene concentration. This is also seen in the data of Table 1. The maximum of excimer emission appears in about the same position as the one observed with pyrene solution in pure solvents (cf. Figure 5a). This is again a n indication that excimer does not come from ground-state dimers which would emit a t shorter wavelengths.l* It is interesting to note that the monomer pyrene emission has the same structure a t low and high pyrene concentration. This means that no self-quenching (by reabsorption due to the fluorescenceabsorption overlap) occurs, even a t high pyrene concen(17) Winnik, F. M. Chem. Rev. 1993,93,587. (18)L'Heureux, G. P.; Fragata, M. J. Photochem. Photobiol., B 1989, 3, 53

4358 Langmuir, Vol. 11, No. 11, 1995 tration. This phenomenon is due to the thinness of the film and we commonly observe it with such films. Pyrene in thin film finds itself in a very polar environment, as seen from the comparative spectra of Figure 5b. By comparing the 11/13ratio of the peak intensities,lg we note that the pyrene environment in the film is more polar than in a solution in methanol. Table 1shows values of pyrene monomer decay times, separated by an interference filter at 380 nm. The decay profiles were always fitted by a sum of two exponentials. The shorter component (32ns), which gave the smaller contribution, was not affected by pyrene concentration. It must be a result of interactions between pyrene and its environment. Such phenomena are frequently encountered with pyrene solubilized in polar environments and, in particular, when pyrene is adsorbed on active surfaces. Note in this respect the works by Liu, de Mayo, and Warez0%21 studying decay time distribution of polycyclic aromatic hydrocarbons on silica gel surfaces. Pankasem and Thomasz2have found, by studying pyrene adsorbed on alumina, that high-temperature dehydroxylation creates favorable conditions for charge transfer complexes which are then responsible for extensive decay time decrease. Pyrene molecules complexedby charge transfer, of course, do not participate in excimer formation. The longer component's lifetime varied (decreased) as pyrene concentration increased. We believe that it is this component that participates in excimer formation and that the decrease of its lifetime is due to quenching. We may then consider the value sz=142 ns, recorded with the lowest pyrene concentration as the decay time of monomer (unreacted) pyrene in the film. This value is much lower than the value obtained for pyrene in solid matrices, e.g., silicate xerogels (see below), which exceeds 300 ns. The lower decay times are obviously due to oxygen quenching. Pyrene then in the Triton film is not incorporated in a matrix but it is accommodated in an environment where oxygen diffusion is possible. It is interesting to note that pyrene decays a t the same rate when adsorbed on low temperature treated (non-dehydroxylated) alumina.22 Figure 6 shows the fluorescence decay profiles of monomer and excimer pyrene (normalized a t maxima). There are some features of excimer emission profile which need special attention. The excimer profile, clearly, demonstrates a rising part. This might indicate that excimer is formed in a dynamic process and it is not due to preassociated pyrene molecules. Note also that the excimer decay profile and the monomer decay profile are parallel a t long times when pyrene concentration was 16 mM. This fact simply means that the excimer decay time is very short so that its fluorescence emission rate is controlled by the rate of excimer formation. Because excimer decays fast, its dissociation rate is then negligible. Of course, the above finding also means that excimer is purely a n excited-state and not a ground-state dimer, in confirmation of previous arguments. As a final confirmation of the fact that pyrene does not form ground-state dimers was the study of the excitation ~ p e c t r a l(not ~,~~ shown), with the emission wavelength fixed a t 385,420, and 480 nm. They were identical within experimental error. The above results can be explained if we accept the following model: titanium isopropoxide is partially hy(19) Lianos, P.; Georghiou, S. Photochem. Photobiol. 1979,30,355. (20) Liu, Y. S.;Ware, W. R. J.Phys. Chem. 1993,97, 5980. (21) Liu, Y. S.; de Mayo, P.; Ware, W. R. J. Phys. Chem. 1993,97, 5987 and 5995. (22)Pankasem, S.; Thomas, J. K. J . Phys. Chem. 1991,95, 7385. (23) Fujii, T.; Shimizu, E.; Suzuki, S. J. Chem. Soc., Furuduy Trans. 1 1988,24,4387.

Papoutsi et al. 16.00

4.00

1 1

1

0.00o 0 0

200.00

400.00

600.00

800 00

loon no

Channel No. Figure 6. Fluorescence decay profiles of pyrene in a Triton film: (-1 monomer decay (concentrationin solution 16 mM); (- - -) monomer decay (concentrationin solution 1mM); (noisy curve) excimer decay (concentrationin solution 16 mM). The monomer curves have been pretreated to substract scattered light and short-componentcontribution. They are convoluted with the excitation profile. drolyzed when solubilized in cyclohexanecontaining Triton and water. When a properly cleaned glass (or quartz) slide is dipped in the solution, the hydrolyzed groups adhere on the glass surface by forming -Ti-0-Sibonds. Nonhydrolyzed isopropyl groups are exposed. Triton also adheres, apparently, by hydrophobic interaction. Pyrene is carried along, possibly, also by hydrophobic attraction. As soon as the mixture is exposed to ambient conditions, cyclohexane evaporates, while atmospheric humidity hydrolyzes the remaining nonhydrolyzed alkoxide groups. The titania film progressively becomes hydrophilic, as expected for the surface of most oxides exposed to ambient at low temperature.161zzThis hydrophobic to hydrophilic evolution forces surfactant molecules to self-organize in order to deal with the unfavorable conditions created by surface tran~formation.'~ Pyrene is then accommodated in the environment provided by the surfactant film, even though incorporation into the Ti02 membrane is also possible. The titania film is porous.12 For this reason it is characterized by increased capacity of adsorption not only for pyrene but also for several other molecules and atomic ions.24 Dynamic excimer formation is obviously facilitated by both diffusion on the surface of titania particles and within the environment provided by the surfactant film. Substantial pyrene excimer formation has also been observed on y-alumina supports (hydroxylatedlZ6However, when pyrene is adsorbed on silica gel, it is found that ground state dimers are responsible for excimer emission.23 HDBAC films demonstrate a similar behavior to Triton films, as far as pyrene excimer formation is concerned. The case ofAOT films, however, was very different. Figure 7 shows clearly that excimer emission is independent of monomer pyrene emission; i.e. it has a smaller lifetime than the monomer lifetime. Quenching by excimer formation is less efficient in AOT films than it is in Triton films, as seen from the evolution of the values of tz in Table 1. These findings are in favor of the possibility that in AOT films excimer mainly comes from aggregated pyrene molecules and it is not diffusion-controlled. Apparently, the structure of AOT films is such that it does (24) Unpublished results. (25) Mao, Y.; Thomas, J. K. Lczngmuir 1992,8, 2501.

Molecular Diffusion in Surfactant Films

Langmuir, Vol. 11, No. 11, 1995 4359 Table 2. Results of the Analysis of Energy Transfer between Pyrene and Coumarin-153or Pyrene and H7HCa acceptor

concentration

f

(mM)b

K1 (lo6 s-l)

KL(lo6 s-l)

Triton Films, Acceptor = Coumarin-153 0.5 1.0 2.0 2.5

r: 110

i

Ow0

-

LI)OCI

7

1

rr

-1

4 0 0 ID

r l Tr

p

GOriClC

0.5 1.0 2.0 2.5 T”rTTTT7-7

BOG00

lGOC’30

crrJrr,pl hc Figure 7. Fluorescence decay profiles of pyrene in an AOT film: (-1 monomer decay (concentration in solution 16 mM); (noisycurve) correspondingexcimer decay. The monomer curve has been pretreated as in Figure 6.

not permit diffusion ofthe fluorophore. This is in contrast to Triton and HDBAC, which, apparently, form fluid structures, possibly aided by the benzyl group, which would favor a liquid crystalline phase. A plausible explanation of these differences might also be based on the fact that the hydrophobic-hydrophilic balance is different in Triton and HDBAC molecules than it is in AOT. This is a n open question which necessitates further investigations, in order to obtain conclusive results. 3.3. Effect of Cosolubilized Polymer on Pyrene Excimer Formation in Triton Films. Poly(viny1 methyl ether) is soluble a t reasonable amount in cyclohexane containing Triton and water, a t concentrations used to make Triton films. We have performed some preliminary experiments where the polymer content was 20 wt % with respect to Triton. The behavior both of the precursor solution and of the film was not affected by the presence of the polymer. The polymer-containing film was transparent and visually uniform. Nevertheless, the solution gels more rapidly in the presence of polymer. For this reason, care was taken to dip the substrate in the solution immediately afier solubilization of the alkoxide. One reason for this faster gelation rate might be that polymer contains additional water which adds to the hydrolysis of alkoxide molecules. Pyrene excimer formation is partially restricted in the presence of polymer. Excimer intensity is thus decreased by about 25% in the presence of polymer. Apparently, the presence of poly(vinyl methyl ether) results in the creation of new solubilization sites for pyrene, “diluting” the fluorophore in the film and decreasing the possibility of excimer formation. This subject is further studied in our laboratory. 3.4. Energy Transfer from Pyrene to Coumarin or to H7HC. Energy transfer studies were made using pyrene as donor and coumarin-153 or H7HC as acceptor. The concentration of pyrene in the original solution was kept a t 2 mM while those of coumarin and H7HC ranged from 0.5 to 2.5 mM and from 1to 5 mM, respectively. The fluorescence decay profiles of the donor in the presence of acceptors have been analyzed by using a model of stretched exponentials, previously employed to study diffision and energy transfer in lipid vesicles.26-28In that model, the (26) Lianos, P.; Duportail, G. Biophys. Chem. 1993,48, 293. (27) Lianos, P.;Argyrakis, P. J. Phys. Chem. 1994, 98, 7278.

1.0 2.0 3.0 5.0

0.53 0.59

77 115 0.54 200 0.59 205 AOT Films, Acceptor = Coumarin-153 0.45 116 0.45 112 0.53 106 0.56 113 AOT Films, Acceptor = H7HC 0.13 1384 0.14 1297 0.21 1321 0.19 1467

0.9 2.5 2.5 4.4

1.1 1.5 2.9 3.5 4.1 4.2 5.6 7.5

a Pyrene concentration in the original solutionwas 2 mM. F’yrene monomer decay time was taken 142 and 159 ns in Triton and AOT films, respectively. Concentrations in the original solutions.

*

fluorescencedecay profile and the first-order rate are given by the following equations

where CI,CZ, and f a r e positive constants. f is always smaller than unity. Equation 1is derived by assuming that the reaction proceeds by random walk in a fractal domain. f then is equal to half of the spectral dimension of the random walk. f is proportional to the fractal dimension and it is smaller in a more restricted environment. C1 is the parameter that describes the efficiency of the reaction; i.e., the value of K(t)is mainly affected by CI. C2 is a parameter ensuing from a second-order approximation to the decay profile27and it mainly depends on the ability of reactants to diffuse within the reaction domain, during the lifetime of the excited state.28 When we study quenching by energy transfer, we find that CZ is equal to zero for immobilized reactants (see below) but C2 is different from zero for mobile reactants. Even though, the derivation of eq 1is based on fractal theory, this is in fact a more general equation27which applies to any restricted environment no matter whether it is fractal or not. Reactant movement then comes as a perturbation27,28to the general equation for energy t r a n ~ f e r . ~ ~ , ~ ~ Table 2 shows the results of the analysis by using the above eqs 1and 2. Data are presented only for Triton and AOT films. HDBAC films were not studied by energy transfer since the analysis of pyrene excimer-formation data revealed a similar behavior as in Triton films. Also, H7HC was studied only in the case ofAOT films since this substance is insoluble in solutions containing nonionic (or cationic) surfactants, at the concentrations used here. We must first stress the fact that CZwas always found different from zero. This means that reactants are mobile, so that reaction configuration changes during the excited state of pyrene (tequals about 150 ns). We have performed some test measurements in transparent solid (28) Duportail, G.; Merola, F.; Lianos, P. J. Photochem. Photobiol., A 1996,89,135. (29)KlaRer, J.;Blumen, A. J. Chem. Phys. 19f34,80,875. (30)Pines,D.; Huppert,D.;Avnir,D. J. Chem. Phys. 1988,89,1177.

4360 Langmuir, Vol. 11, No. 11, 1995

matrices made by hydrolysis of tetrameth~xysilane,~~ using again pyrene as a donor and coumarin or H7HC as acceptor. Analysis of the fluorescence decay profiles with the help of eq 2 in such matrices results in Cz = 0, in all studied cases. This confirmed the above assertion that motion of donor and acceptor molecules comes as a perturbation (second-order effect) to the general energy transfer scheme.30 As seen in Table 2, the values off obtained with coumarin are about the same in the two films, with a small tendency to diminish in the case of AOT. f for coumarin is close to 0.5, a value that corresponds to randomly distributed reactant^.^^^^^ This result indicates that both films provide a rather non restricted space for coumarin distribution. The situation is very different in the case of H7HC. fvalues were particularly small with this acceptor. This result suggests that pyrene and H7HC find themselves solubilized in a very-low-dimensional structure, when they are in AOT films. The finding ofthe previous paragraph, that pyrene has a tendency to aggregate in AOT, suggests that molecular motion is restricted for some molecular species in the films of this surfactant. This fact is combined with the tendency demonstrated by amphiphilic hemicyanine derivatives, such as H7HC, to form aggregates when they are deposited in pure or mixed monolayer^.^^ Mention should be also made in this respect to the work of Ashwell et al.,33 suggesting a special structure obtained by combination of H18HC with an anionic surfactant (octadecyl sulfate). Aggregation of H7HC in AOT films whould correspond to low-dimensional structures yielding lowfvalues. However, further information listed in Table 2 makes this question more complicated. The reaction rates for energy transfer between pyrene and H7HC are much higher than in the case of the pyrene-coumarin couple. The reaction rates are shown in the last two columns of Table 2 as K1 and& K1 represents the reaction rate a t the first recorded time channel, giving the reaction (energy transfer) efficiency a t short times. KL is the reaction rate a t the last recorded time channel (i.e. 660 ns after the beginning of the excitation pulse). KL represents the reaction probability at long times. However, we note that KL for H7HC has a substantial value. This is against the possibility of reactant aggregation, since the reaction rate for aggregated species should be zero a t long times. Aggregation of H7HC should be obvious in the absorption spectrum of this dye. Indeed, a t H7CH concentrations substantially higher than the ones used here, aggregation is demonstrated by an additional absorption peak at around 400 nm (not shown here).24 Such a peak is not detected in the spectrum of Figure 3b. Apparently this question cannot be clearly answered by the data presented here. For this reason the problem is further studied in our laboratories. It is nevertheless (31) Modes, S.; Lianos, P. Chem. Phys. Lett. 1988,153, 351. (32) Hall, R. A,; Thistlethwaite, P. J.; Grieser, F.;Kimizuka, N.; Kunitake, T.J.Phys. Chem. 1993,97,11974. (33) Ashwell, G. J.; Hargreaves, R. C.; Baldwin, C. E.; Bahra, G. S.; Brown, C. R. Nature 1992,357,393.

Papoutsi et al. obvious that AOT film structure is not similar to the one obtained with Triton or HDBAC which seem to produce more uniform liquid-like structures, while AOT yields more restricted structures. 3.5. Location of Probes in the Films. Even though the above data are not suficient to draw a clear picture about the location of the probes in the films we make the following suggestions. H7HC most probably follows the geometry of AOT in the film because of its amphiphilic nature and the length of its aliphatic chain. Pyrene should be located close to the polar interface between titania and surfactant. This assumption is justified by the fact that the 11/13 intensity ratio in its fluorescence spectrum suggests a very polar environment, as already said. Apparently, the complete hydrolysis of nonhydrolyzed alkoxide groups occurs rapidly upon exposure to atmospheric humidity so that the final situation is obtained right after withdrawal of the slide from the solution. This is justified by the finding that the spectroscopic characteristics of the films (bearing probes) do not substantially vary with time. The films are three-dimensional, in the sense that they are much thicker than the size of pyrene. Still pyrene is preferentially solubilized a t the interface which makes its distribution more planar. The value of the decay time of pyrene, as already said, is supportive to such an assumption. Coumarin is slightly soluble in water and very soluble in cyclohexane. Its hydophilicityhydrophobicity is not very different from that of pyrene. We believe that it behaves in a similar way as pyrene; i.e., it also preferentially resides a t the interface. Even though the exact geometry of pyrene and coumarin distribution cannot be accurately determined by the present data, it is obvious that coumarin possesses a large degree of freedom; Le., it is randomly distributed at the interface, contrary to H7HC (and to excimer-formingpyrene in AOT) which tend to form aggregates. 4. Conclusions The kinetics of pyrene excimer formation reveals that molecular diffusion is possible in Triton and HDBAC films which, possibly, form liquid crystalline structures. On the contrary, AOT films provide a rigid environment. Pyrene aggregates when it is in AOT films. The presence ofpolymer chains in Triton films creates new solubilization sites for pyrene and decreases the excimer forming efficiency. H7HC can be incorporated only in AOT films, apparently by electrostatic attraction. H7HC appears to form specific low-dimensional domains in AOT films. Triton and HDBAC films appear to have similar properties as far as molecular motion inside them is concerned. AOT demonstrates a different behavior. Further investigations are necessary to obtain a clearer picture of the system.

Acknowledgment. We acknowledgefinancial support from the Human Capital and Mobility Project CHRX-CT0273 and from the FIENEA Project 91 EA 815. LA9502352