Compartmentalization of Reactants in Different Regions of Sodium 1,4-Bis(2-ethylhexyl)sulfosuccinate/Heptane/Water Reverse Micelles and Its Influence on Bimolecular Electron-Transfer

Compartmentalization of Reactants in Different Regions of Sodium1,4-Bis(2-ethylhexyl)sulfosuccinate/Heptane/Water Reverse Micelles and Its Influence onBimolecular Electron-Transfer Kinetics

Sharmistha Dutta Choudhury, Manoj Kumbhakar, Sukhendu Nath, Sisir Kumar Sarkar,Tulsi Mukherjee, and Haridas Pal*

Radiation and Photochemistry DiVision, Bhabha Atomic Research Centre, Mumbai 400 085, India

ReceiVed: March 20, 2007; In Final Form: May 11, 2007

Sodium 1,4-bis(2-ethylhexyl)sulfosuccinate (AOT) reverse micellar medium has been used to study thephotoinduced electron-transfer (ET) reactions between some coumarin derivatives and amines, namely, aniline(AN) andN,N-dimethylaniline (DMAN) at differentw0 (w0 ) [water]/[AOT]) values, to explore the appearanceof Marcus inversion and also the possible role ofw0, if any, on the Marcus correlation curves. The coumarinderivatives are found to partition between the heptane-like and the water-like phases of the reverse micelles,and their locations have been confirmed by time-resolved anisotropy measurements. Fluorescence quenchingis found to depend both on the location of the coumarin molecules and on the hydrophobicity of the aminedonors. Various aspects such as the effect of differential partitioning of the quenchers, the location of theprobes in the two phases, the diffusion of the reactants in the micellar phase, etc. have been considered torationalize the fluorescence quenching rates in reverse micelles. Rotational relaxation times and the diffusionparameters estimated from the anisotropy results do not show good correlation with the observed quenchingrates indicating that the diffusion of reactants has no role in the quenching kinetics in reverse micelles. Marcusinversion behavior has been observed for the coumarin-amine systems in the water-like phase at a relativelyhigh exergonicity of∼1.2 eV suggesting that the solvent reorganization energy contributes fully to the freeenergy of activation for the ET reactions in the present systems. This is in accordance with the fast solventrelaxation dynamics reported in reverse micelles. Quenching rates in the water-like phase are found to decreaseor increase marginally with increasingw0 for the coumarin-DMAN and coumarin-AN systems, respectively.This is explained on the basis of the changing solubility of these amines in the water-like phase with changingw0 values of the reverse micelles. In the heptane-like phase, no clear inversion in the quenching rate versusfree energy plot could be observed because the study could not be extended to higher exergonicity due tononsolubility of the dye C151 in this phase. Present results, especially in the water-like phase, suggest thatthe confinement of reactants in micellar media can effectively remove the influence of reactant diffusion onbimolecular ET rates and thus make the systems more conducive for the observation of the Marcus invertedregion.

1. Introduction

Reverse micelles continue to be the subject of intensiveinterdisciplinary research. In addition to providing a suitablereaction medium for species that are insoluble in nonpolarsolvents, reverse micelles serve as templates for nanomaterialsynthesis and are also useful as models for biologicalcompartmentalization.1-5 In view of the ease of preparation,monodispersity, stability, and ability to solubilize large amountsof water, sodium 1,4-bis(2-ethylhexyl)sulfosuccinate (AOT) isthe most widely used and well characterized surfactant in thestudy of reverse micelles. Structurally, the AOT/heptane/waterreverse micelles consist of an inner water pool that is surroundedby a layer of surfactant molecules with their polar head groupsfacing the water pool and the nonpolar tails projected outsidetoward bulk heptane (Scheme 1). The water pool along withthe hydrated surfactant head groups form the typical polar water-like phase, and the surfactant tails along with the intervening

heptane molecules form the typical nonpolar heptane-like phasein the form of a shell encircling the water pool. The radius ofthe water pool inside AOT reverse micelles is found to beroughly 2w0 Å, wherew0 is defined as the [water]/[AOT] ratio.3

Accordingly, the size of reverse micelles, especially the waterpool size, can be easily manipulated by varying the water tosurfactant molar ratio, and consequently the properties such aspolarity, viscosity, etc. of the water-like phase can also bevaried.6-8 Infrared, NMR, and other spectroscopic techniqueshave revealed many interesting properties of the water moleculespresent inside AOT reverse micelles.9-17 It has been shown thatthree different types of water molecules are present in reversemicelles, namely, the free water, the bound water, and thetrapped water.9 The free water in reverse micelles is quite similarto the bulk state of water, containing hydrogen-bonded chainsof water molecules. The contribution of these water moleculesincreases profoundly as thew0 value of the micelle increases.The bound water in reverse micelles represents those watermolecules that are hydrogen-bonded to the surfactant head

* Author to whom correspondence should be addressed. Fax: 91-22-25505151/25519613. E-mail: [email protected].

8842 J. Phys. Chem. B2007,111,8842-8853

10.1021/jp0722004 CCC: $37.00 © 2007 American Chemical SocietyPublished on Web 07/04/2007

groups while the trapped water refers to the matrix-isolated watermolecules, mainly monomeric or dimeric, which are trappedbetween the surfactant chains near the interface. The boundwater and trapped water are mainly responsible for the hydrationof the surfactant head groups at the interfacial region.

Among many different types of chemical reactions that havebeen investigated in reverse micelles, the photoinduced electron-transfer (ET) reactions deserve special attention.18-24 Restrictiveenvironments are known to influence ET reactions favorablyfor their various applications, and the use of reverse micelleshave proved to be quite useful in these respects.18 Studies haveshown that reaction rates are greatly altered in reverse micellescompared to those in homogeneous media, and these changesare also dependent on the size of the water pools. It has beenfound that the fraction of free water in the water pool oftenplays an important role in governing the efficiency of ETreactions.20 The localization sites and the orientations of thesubstrates as well as the sign of the charges at the interfacialregions are also found to be crucial factors in determining theefficiency of many reactions in reverse micelles.19,24

In our group, substantial work has been carried out onphotoinduced ET reactions in normal micelles, where the mostimportant result is the observation of the “inverted region” inthe Marcus correlation of the bimolecular ET rates with the freeenergy changes of the ET reactions.25-27 According to the

classical Marcus theory for outer-sphere ET reactions, the rateconstant,kET, is given by

whereν is the frequency of motion in the reactant potentialwell, ∆G° is the free energy change for the ET reaction,kB isthe Boltzmann constant,T is the absolute temperature, andλ isthe total reorganization energy, given asλ ) (λs + λi), whereλs and λi are the solvent reorganization energy and theintramolecular reorganization energy, respectively.28,29The mostinteresting feature of this theory is the expected inversion inthe ET rates as the exergonicity of the reaction (-∆G°) exceedsλ. Although predicted in 1956, the inverted region had baffledresearchers for a long time, and the first experimental demon-stration of the inverted region was made only in the mid-1980sfor intramolecular ET reactions.30-31 Though many intramo-lecular ET reactions and charge recombination reactions inradical ion pairs are now known to show inversion behavior,for intermolecular or bimolecular ET reactions, the invertedregion is still quite elusive, as in most of these cases the ETrates are limited by the diffusional rates of the reactants.28-29

This problem of diffusion in bimolecular ET reactions canhowever be circumvented by using micellar media where

SCHEME 1: Molecular Structures of AOT, the Coumarin Derivatives, and the Amines

kET ) ν exp(-(∆G° + λ)2

4λkBT ) (1)

Reactants in AOT/Heptane/Water Reverse Micelles J. Phys. Chem. B, Vol. 111, No. 30, 20078843

diffusion of the reactants is largely retarded or prevented, andthus the actual ET rates can be monitored.25-27 Moreover, inmicelles, due to slow solvation dynamics, the effective solventreorganization energy (λs) for the ET reaction can be lower thanthat in homogeneous solvents with fast solvation dynamics.25-27

This lowering in effectiveλs will shift the inverted region inthe Marcus Correlation (ket vs ∆G° plot) to somewhat lowerexergonicity and thus make it easy to be observed. In thiscontext, the study of ET processes in reverse micelles is veryappealing, especially to understand whether a situation similarto normal micelles also prevails in reverse micelles in relationto the observation of Marcus inversion behavior for bimolecularET reactions. In the present work we have investigated in detailthe steady-state and time-resolved fluorescence quenching ofdifferent coumarin derivatives by amine donors such as aniline(AN) andN,N-dimethylaniline (DMAN) in AOT/heptane/waterreverse micelles to understand the effect of the topology of thereverse micelle on bimolecular ET reactions As the coumarindyes can solubilize in both the water-like and the heptane-likephases of the reverse micelles, measurements have been carriedout at suitable wavelengths to probe the dyes predominantly ina particular phase. Different locations of the coumarin dyes inthe reverse micelles have been characterized by using absorptionand fluorescence spectral characteristics as well as fluorescenceanisotropy decay measurements. The amine quenchers have beenchosen on the basis of their widely different water/heptanepartition coefficients (Kpartition )1 and 1/251 for AN and DMAN,respectively)32 so that they will have different preferentialsolubility in the water-like and heptane-like phases.

2. Materials and Methods

Laser grade coumarin dyes were obtained from Exciton andused as received. AN and DMAN were obtained from Spec-trochem (India) and were purified by vacuum distillation justbefore use. AOT was obtained from John Baker and was driedin vacuum prior to use. Heptane from Spectrochem (India) wasused without further purification. Nanopure water was obtainedby passing distilled water through a Barnstead Nanopure WaterSystem and was systematically added to prepare reverse micellesolutions with desiredw0 values. The AOT concentration waskept 0.2 M throughout. The total coumarin concentrations werealways kept much lower (∼10-20 × 10-6 mol dm-3) than thereverse micelle concentrations (∼0.2-1.7 × 10-3 mol dm-3),at all w0, so that no more than one dye molecule could occupyan individual micelle. Freshly prepared solutions were usedthroughout. The chemical structures of the acceptors (coumarindyes), the donors (AN and DMAN), and AOT used in this studyare given in Scheme 1 along with their abbreviations.

A Shimadzu UV-vis spectrophotometer (model UV-160A)was used for recording the absorption spectra, and a Hitachispectrofluorometer (F-4010) was used for measuring the steady-state fluorescence spectra. An IBH instrument, which is basedon the time-correlated single photon counting (TCSPC) prin-ciple, was used for the time-resolved fluorescence measure-ments. A picosecond diode laser (NanoLed-07, 408 nm) wasused as the excitation source, and a photomultiplier-baseddetection module (TBX-04) was used for measurements of thefluorescence decays. Except for anisotropy measurements, allother fluorescence decays were collected at the magic angle(54.7°) with respect to the vertically polarized excitation lightto avoid the effect of rotational relaxation of the dyes on theirfluorescence decays. The instrument response function for thepresent setup is∼230 ps (full width at half-maximum (fwhm)).DAS-6 software obtained from IBH was used for the decon-

volution analysis of the observed decays, considering eithermonoexponential or biexponential decay functions.

For fluorescence anisotropy measurements, the polarizedfluorescence decays,I|(t) andI⊥(t), whereI|(t) andI⊥(t) are thedecays for the parallel and perpendicular emission polarizationswith respect to the vertical excitation polarization, were firstcollected. Using theseI|(t) andI⊥(t) decays, the anisotropy decayfunction r(t) was constructed as33

whereG is a correction factor for the polarization bias of thedetection setup. TheG factor was obtained independently bymeasuring the two perpendicularly polarized fluorescence decaysand using horizontally polarized light for sample excitation.33

These measurements were performed in duplicate to check thereproducibility and to obtain the average values of the relaxationtimes.

As will be discussed later, most of the coumarins and aminesused in the present study are solubilized in both the water-likeand the heptane-like phases of the reverse micelles. Due to thisreason, their redox potentials could not be estimated unambigu-ously using cyclic voltammetric (CV) methods. Unlike the othercoumarins, the dye C151 is seen to solubilize almost exclusivelyin the water-like phase. Thus, the reduction potential of C151in AOT reverse micelles was measured using CV methods withan Eco-Chemie Potentiostat/Galvanostat-20 coupled with GPESsoftware. A solution of C151 in AOT reverse micelles contain-ing tetraethyl ammonium bromide as the supporting electrolytewas first deaerated by purging high-purity N2 gas for about30 min. The CV measurements were then carried out usingdropping mercury as the working electrode, a graphite rod asthe counter electrode, and a standard calomel electrode as thereference electrode. On the basis of this value, the redoxpotentials for the other coumarin dyes and amine donors in thewater-like phase of the reverse micelles were estimated byapplying a suitable correction to their reported redox potentialsin acetonitrile solutions.34-37 As the reduction potential of C151does not change any significantly withw0, it is assumed thatfor the other reactants also the redox potentials remain invariantwith w0. For the heptane-like phase, the redox potentials of thereactants could not be measured in the present study.

3. Results and Discussion

3.1. Ground-State Absorption Measurements.Absorptionspectra of the coumarin dyes were measured in heptane, water,and AOT reverse micelles atw0 ) 10, 20, and 40. For C151,the absorption spectrum in AOT reverse micelles resembles thatin water indicating that these molecules solubilize predominantlyin the water-like phase of the reverse micelles. This is quiteexpected as due to the presence of the amino hydrogens C151can easily undergo intermolecular hydrogen-bonding interactionswith the available water molecules and also with the surfactanthead groups. For all other coumarins, the absorption spectra inAOT are similar to that in heptane, but the spectra are in generalbroad and show a shoulder at the longer wavelength regioncorresponding to the spectra in water. Representative spectrafor C522 are shown in Figure 1. It is thus indicated that thesedyes are solubilized more in the heptane-like phase than in thewater-like phase of the reverse micelles. As the solubility ofall the coumarin dyes is very low in pure water, it is unlikelythat these dyes will reside deep inside the water pool of thereverse micelles. We thus expect that in the water-like phase

r(t) )I|(t) - GI⊥(t)

I|(t) + 2GI⊥(t)(2)

8844 J. Phys. Chem. B, Vol. 111, No. 30, 2007 Choudhury et al.

the dye molecules will preferentially reside at the surfactant-water interfacial region constituted by the surfactant head groupsand the bound and trapped water molecules (cf. Scheme 1). Thisis further supported by the observation that the absorptionspectra of the coumarin dyes remain almost unchanged onchanging thew0 values of the reverse micelles. This is expectedsince the polarity at the interfacial region does not changesignificantly with an increase inw0 although the characteristicsof the free water inside the water pool become more like thoseof bulk water. The dyes in the heptane-like phase could bepresent either in the bulk heptane outside the reverse micellesor in the nonpolar shell of the reverse micelles constituted bythe surfactant tails and intervening heptane molecules. Sincethe solubility of the coumarin dyes in a nonpolar hydrocarbonsolvent is also quite low, it is possible that the dyes in theheptane-like phase actually prefer to reside in the nonpolarshell, which can also contain some dispersed water moleculesas this shell is adjacent to the strongly hydrated head groupregion. Further support for location of the dyes in the nonpolarshell is also obtained from time-resolved fluorescence studiesas will be discussed later. However, the solubilization of a smallfraction of the dyes in bulk heptane cannot be ruled outcompletely.

It is seen that the absorption spectra of the coumarin dyes inAOT reverse micelles remain unchanged on addition of theamines (AN and DMAN) in the solution. This indicates thatthere is no ground-state complex formation between thecoumarins and the amines in the reverse micelles. On the basisof the water/heptane partition coefficients of the amines used(Kpartition )1 and 1/251 for AN and DMAN, respectively),32 itis expected that these amines will also be solubilized in differentregions of the reverse micelles in varying proportions. Thus, inthe water-like phase, AN will be much more soluble thanDMAN where as in the heptane-like phase the situation will bereversed.

3.2. Fluorescence Spectral Measurements.Figures 2a and2b show representative emission spectra for C151 and C522dyes, respectively, in heptane, water, and AOT reverse micelles.The emission spectrum for C151 in reverse micelles resemblesthe spectrum in water, again indicating that this dye solubilizespredominantly in the water-like phase. However, the emissionspectrum of C151 in AOT is significantly blue-shifted comparedto the spectrum in aqueous solution. This possibly indicates thatthe dye molecules in the water-like phase are actually at thesurfactant-water interfacial region rather than deep inside thewater pool. The emission spectrum for C522, as shown inFigure 2b, displays a clear signature for the presence of thedye molecules in both the heptane-like phase (emission maxima∼440 nm) and the water-like phase (emission maxima

∼512 nm). Substantial blue shift of the emission peak for thedye C522 in the water-like phase compared to that in pure wateris the indication for the presence of the dye molecules at thesurfactant-water interfacial region as mentioned previously forC151. For the other coumarin dyes, the emission spectra in AOTreverse micelles are qualitatively very similar to that of C522.Thus, the emission spectra of these dyes show the band for theheptane-like phase around 420-440 nm and the band for thewater-like phase around 500-530 nm. The emission maximaof the coumarin dyes in heptane, water, and the heptane-likeand water-like phases of AOT reverse micelles are listed inTable 1.

3.3. Time-Resolved Anisotropy Measurements.To gainmore insight on the localization sites of the dyes in AOT reversemicelles, the fluorescence anisotropy decays of the coumarindyes were measured both around 420-440 nm for the dyes inthe heptane-like phase and around 500-530 nm for the dyes inthe water-like phase. In the heptane-like phase, the anisotropydecays for all of the dyes were found to be single-exponential.The anisotropy decays, however, showed biexponential behaviorin the water-like phase. Anisotropy decays along with their fittedcurves for C522 in the two phases are shown in Figure 3. For

Figure 1. Normalized absorption spectra of C522 in heptane (---),water (‚‚‚), and 0.2 M AOT reverse micellar solution (s) at w0 ) 20.

Figure 2. Steady-state fluorescence spectra of (a) C151 and (b) C522in heptane (---), water (‚‚‚), and 0.2 M AOT reverse micellar solution(s) at w0 ) 20 and fluorescence quenching in the reverse micellarmedia with (1) 0, (2) 0.006, (3) 0.02, (4) 0.04, (5) 0.06, and (6)0.1 dm3 mol-1 DMAN.

TABLE 1: Emission Maxima (λmax) of the CoumarinDerivatives in Heptane, Water, and AOT Reverse Micelles

λmax (nm)

AOT

coumarin heptane water heptane-like water-like

C153 451 547 454 526C522 440 530 440 512C152 426 530 429 513C151 403 495 490


C151, the anisotropy decay in the heptane-like phase could notbe measured as the dye is almost exclusively solubilized in thewater-like phase. The rotational relaxation times (τrot for single-exponential,τfast andτslow for biexponential) and their relativecontributions (afast andaslow) as estimated from the analysis ofthe anisotropy decays are listed in Table 2.

The τrot values for the dyes in the heptane-like phase are inthe range of 200-300 ps. In pure heptane,τrot values for thecoumarin dyes are about 30 ps.38 So in the heptane-like phaseof the reverse micelles, the dyes experience much more viscousfriction than in pure heptane. This is expected if the dyes in theheptane-like phase actually reside in the nonpolar shell of thereverse micelles where the microviscosity will be much higherthan that in bulk heptane. The single-exponential nature of theanisotropy decays in the heptane-like phase, however, indicatesthat the microenvironment in the nonpolar shell is quite isotropicin nature.

Biexponential anisotropy decay in the 500-530 nm regionis a strong indication that the dyes in the water-like phase residein the interfacial region and are intimately associated with thesurfactant head groups. As a result, the rotation of the dyemolecules will be quite anisotropic in nature, and the anisotropydecays will be non-single-exponential, as reported in manyearlier studies in micelles and reverse micelles.39-41 It is seenfrom Table 2 that the contribution,afast, for the fluorescenceanisotropy decays of the dyes in the water-like phase increasesmoderately with increasingw0. This we attribute to the effectof increased hydration of the surfactant head groups withincreasingw0. Interestingly, however, theτfast andτslow valuesdo not change significantly withw0. As the dyes in the interfacialregion are strongly associated with the surfactants, a marginalincrease in the hydration of the head groups cannot cause anyprofound effect on the rotation of these dye molecules.

Fluorescence anisotropy decays of the dyes in the 500-530 nm region can be analyzed using the two-step model asapplied earlier in many anisotropy studies in micellar media.38-40

According to this model, three different kinds of motions cancontribute to the fluorescence anisotropy decay in micelles. Theyare: (i) wobbling motion of the dye, (ii) lateral diffusion of thedye on a hypothetical spherical surface in the micelle, and (iii)rotation of the whole micelle in the solution. Consideringτw,τL, andτM as the respective correlation times for the above threemotions, the observedτfast and τslow values can be expressedas39-41

As theτfast andτslow values do not change much withw0, theestimatedτw values also remain similar at all of thew0 values(cf. Table 2). According to the two-step model, the relativeamplitude aslow of the slow anisotropy component carriesinformation about the restriction experienced by the probe forits rotation in the micellar phase. This is expressed in terms ofan order parameterS, where

The S values estimated for the present systems are listed inTable 2. It is seen that theSvalues are always quite high at allof thew0 values, which provides further evidence that the dyesare localized at the interfacial region where they are stronglyentangled with the surfactant head groups. The slight decreasein the S values on increasingw0 suggests that the microenvi-ronment in this region gradually becomes more relaxed due torelatively loose packing of the surfactants at higherw0 values.Using theseS values, it is possible to calculate the diffusioncoefficientDw for the wobbling motion of the dye as39-41

wherex ) cosθ andθ is the semiangle of the cone in whichthe dye undergoes its wobbling motion in the interfacial region.This angle is related toS by eq 7

The Dw values estimated for the present systems at differentw0 values are listed in Table 2. It is seen thatDw increases withw0 as expected due to an increase in the hydration of thesurfactant head groups at higherw0 values.

Following the Debye-Stokes-Einstein relation, the correla-tion time τM for the rotation of the whole micelle can beexpressed as

whereη is the viscosity of bulk heptane outside the reversemicelle,kB is Boltzmann’s constant,T is the absolute temper-ature, andrM is the radius of the reverse micelle. In the presentcases, theη value is considered to be similar to that of pureheptane, andrM is estimated as the sum of the radiusrwp of thewater pool and the lengthl of the surfactant chain, as found inthe literature.19,20 The τM values thus obtained for the presentsystems are about 9.5, 29.5, and 163.7 ns, respectively, atw0 ) 10, 20, and 40. As theseτM values are much higher thanthe τslow values (cf. Table 2), it is evident from eq 4 that therotation of the whole micelle hardly contributes to the observedanisotropy decays. Thus, we can assume that for the presentsystemsτL ≈ τslow. Sinceτslow does not vary withw0, it impliesthatτL also remains invariant withw0. Such a result is, however,quite improbable, becauserwp increases very sharply withw0,and consequently the dye will have to undergo unusually fastlateral diffusion at higherw0 values to keepτL unchanged.Consideringrwp as the radius of the spherical surface on which

Figure 3. Fluorescence anisotropy decays,r(t), for C522 in AOTreverse micelle (w0 ) 40) at (1) 440 nm and (2) 510 nm. The solidlines show the best fit curves.

τfast-1 ) τw

-1 + τslow-1 (3)

τslow-1 ) τL

-1 + τM-1 (4)

S) xaslow (5)

Dw ) 1

[(1 - S2)τw][x2(1 + x)2

2(x - 1) {ln(1 + x2 ) +

(1 - x)2 } +

(1 - x)24

(6 + 8x - x2 - 12x3 - 7x4)] (6)

S) (cosθ + cos2 θ)/2 (7)

τM )4πrM

3η3kBT

(8)


the dye undergoes the lateral diffusion, the correspondingdiffusion coefficientDL can be expressed as39,40

TheDL values thus estimated for the present systems are listedin Table 2. It is seen that theDL values increase extraordinarilywith increasingw0. This is very unrealistic because the lateraldiffusion of the dyes entangled with the surfactant head groupsis not expected to increase that significantly even if the hydrationof the surfactant head groups increases to some extent. A similarsharp increase in the estimatedDL values was also encounteredby De Schryver and co-workers at higherw0 values in reversemicelles.42 They rationalized this observation by proposing thatthe τslow component of the fluorescence anisotropy decays inreverse micelles is not only a function of the lateral diffusionof the dye but also a function of the breathing motion orsegmental motion of the micelle. As the breathing motion causesa deformation of the reverse micelle, it effectively modifies theτslow component of the anisotropy decay. According to DeSchryver and co-workers, the contribution of this breathingmotion on theτslow component increases very sharply forw0 g10. We believe that such an effect is also predominant in thepresent systems and accordingly the estimation of theDL valuesbased on eq 9 is not justified. We are however not in a positionto quantify this effect and estimate the actual lateral diffusioncoefficients for the present systems. Another useful observationfrom the anisotropy studies is that the correlation times fordifferent coumarin dyes are almost in the same range at anyparticularw0 value. Hence, it can be assumed that the microen-vironment in the heptane-like phase or in the water-like phaseremains more or less similar for all of the dyes.

3.4. Fluorescence Quenching Measurements.As the cou-marin dyes are solubilized in both the water-like and theheptane-like phases, the fluorescence quenching was monitoredin the 420-440 nm region (for the dyes in the heptane-likephase) as well as in the 500-530 nm region (for the dyes inthe water-like phase). From steady-state (SS) measurements,significant quenching of coumarin fluorescence was observedfor both of the phases of the reverse micelles on addition of theamines (Figures 2a and 2b). Drawing inference from previousstudies, it can be stated that the fluorescence quenching in thepresent systems is due to photoinduced ET from the aminedonors to the excited coumarin dyes.25-27 Though the 420-440 and 500-530 nm emission bands of the coumarin dyes

undergo different extents of quenching in the presence of theamines, it has been verified by subtraction of the fluorescencespectra of the dyes in neat heptane from that in AOT reversemicelles that the nature of the difference spectra remainseffectively unchanged for all of the amine concentrations used.This observation indicates that the two emission bands cor-respond to different localization sites of the dyes, and in boththe cases there is no exciplex formation with the amines.

The Stern-Volmer (SV) plots obtained from SS fluorescencequenching studies showed deviations from linearity for mostof the coumarin-amine pairs. Positive deviations were observedfor quenching by AN in both the heptane-like and the water-like phases. Negative deviations were observed for quenchingby DMAN, mostly in the water-like phase. Figure 4 shows somerepresentative SV plots for the quenching of C153 fluorescenceby the two amines atw0 ) 20. Following our earlier studies innormal micelles, the positive deviations can be attributed to high,localized concentrations of the amines (AN) at the solubilizationsites of the coumarins dyes. This can lead to a significant staticquenching of coumarin fluorescence, even in the absence ofground-state complex formation.25-27 Negative deviations canarise either due to the heterogeneous population of fluorophores,each having different accessibility for quencher molecules, ordue to a saturation effect of the quenchers at the dye solubili-zation sites.33,43 Though the former cannot be ruled outcompletely, we feel that the negative deviations in the presentcases are primarily due to the saturation effect as explainedbelow. The coumarin dyes in the water-like phase are locatedin the interfacial region where the micropolarity is considerablyhigh, and thus the solubility of the hydrophobic quencher,

TABLE 2: Time Constants for the Anisotropy Decays as Obtained from Single-Exponential Fits for the Heptane-like andBiexponential Fits for the Water-like Phases along with the Relative Amplitudes for the Latter Casea

heptane-like water-like

coumarin w0

τrot

(ns) afast

τfast

(ns) aslow

τslow ) τL

(ns)τw

(ns)θ

(deg) SDw × 107

(s-1)DL × 10-10

(m2 s-1)

C153 10 0.18 0.29 0.25 0.71 1.16 0.33 26.75 0.85 18.26 4.9720 0.18 0.36 0.29 0.64 1.13 0.38 30.11 0.81 19.29 14.9640 0.18 0.39 0.29 0.61 1.21 0.38 32.0 0.78 21.97 56.53

C522 10 0.22 0.22 0.27 0.78 1.25 0.35 23.21 0.88 12.92 4.6120 0.22 0.27 0.29 0.73 1.38 0.38 25.71 0.86 14.54 12.2540 0.21 0.36 0.31 0.64 1.41 0.39 30.68 0.80 19.01 48.85

C152 10 0.29 0.26 0.39 0.74 1.24 0.58 25.31 0.86 9.27 4.6520 0.27 0.34 0.38 0.66 1.14 0.56 29.27 0.81 12.79 14.8940 0.25 0.38 0.34 0.62 1.09 0.60 31.90 0.79 13.68 62.65

C151 10 0.20 0.39 0.80 1.43 0.53 21.72 0.90 7.58 4.0320 0.24 0.43 0.76 1.55 0.59 24.22 0.87 8.33 10.9040 0.34 0.41 0.66 1.46 0.56 29.54 0.81 12.68 46.85

a Various parameters (τw, θ, S, Dw, τL, andDL), calculated according to the two-step model are also given (see text).

DL )rwp

2

6τL≈ rwp

2

6τslow(9)

Figure 4. Stern-Volmer plots for the steady-state fluorescencequenching of C153 by AN (4) and DMAN (O) in AOT reverse micellesmonitored at 530 nm forw0 ) 20.


DMAN, is expected to be quite low in this region. So as moreDMAN is added to the solution, a saturation effect is possiblefor its solubility in the water-like phase leading to negativedeviation in the SV plots. Since AN is more hydrophilic innature, it can solubilize substantially both in the surfactant headgroup region and in the nonpolar shell of the reverse micelles.So no negative deviation is observed in this case. Instead, thepreferential accumulation of AN in the above-mentioned regionscan cause significant static quenching at higher quencherconcentrations leading to positive deviations of the SV plots. Itis interestingly seen that for all of the coumarin-amine systemsthe nature of the SV plots remains almost unchanged onchanging thew0 values of the reverse micelles. This indicatesthat the solubilization sites of the fluorophores and quenchersare not altered much on changing thew0 values. It is evidentthat the overall quenching process in the present systems is quitecomplex, and to comprehend the SV plots properly, one mustconsider the relative contribution of static quenching as well asthe actual quencher concentrations in the two phases of thereverse micelles. Since it was not possible to estimate theseparameters, we avoided using the SS results and relied solelyon time-resolved (TR) fluorescence measurements for estima-tion of the bimolecular quenching constants,kq, for thesesystems.

For TR studies too, fluorescence decays of the dyes weremeasured for both the heptane-like (420-440 nm) and the water-like (500-530 nm) phases. The decays are always seen to bebiexponential in nature, even in the absence of the amines. Inthe 500-530 nm region, the major component (>95%) of thedecay has a longer lifetime and is attributed to the dyessolubilized in the water-like phase. The minor short lifetimecomponent (<5%) in these decays is possibly due to theoverlapping emission from the dyes dissolved in the heptane-like phase of the reverse micelles. In the 420-440 nm region,similarly, the major decay component (>95%) has a shorterlifetime, corresponding to the dyes in the heptane-like phase,and the minor decay component (<5%) has a longer lifetime,corresponding to the overlapping emission of the dyes in thewater-like phase. In the presence of the amines, the decaysgradually become faster, but their nature remains similar.Representative fluorescence decays for C522 at 440 nm for theheptane-like phase and at 510 nm for the water-like phase areshown in Figures 5a and 5b, respectively, both in the absenceand in the presence of DMAN as the quencher. In the presentcontext it should be mentioned that in our earlier studies in othermicellar media it was observed that in the presence of the aminequenchers the fluorescence decays of the coumarin dyes becomebiexponential in nature, even though the decays were single-exponential in the absence of the amine quenchers.25-27 Asimilar situation is also expected for different coumarin-aminesystems in reverse micelles. In fact, the biexponentiality is seento be more prominent for all of the coumarin-amine systemsin either of the phases of the reverse micelle as the amineconcentrations are increased in the solution (cf. Figure 5). Underthis situation, in the presence of the quenchers, it was notpossible to assign each component of the lifetimes to differentphases of the reverse micelle, though such an assignment ispossible in the absence of the quenchers. Thus, in the presentstudy, we cannot consider the reduction of individual lifetimecomponents separately to estimate quenching constants. Ac-cordingly, the decays were always analyzed as a biexponentialfunction, irrespective of the presence or absence of the quench-ers, and the average lifetime (τav) of the dyes, as defined by

eq 10, was used to determine the quenching constants usingthe SV relation

whereτ1 andτ2 are the two fluorescence lifetimes andA1 andA2 are their relative contributions. Theτav values of the coumarindyes in the two phases in absence of the quenchers arerepresented in general by the symbolτ0 and are listed inTable 3. It is seen thatτ0 increases to a small extent withincreasingw0 in the heptane-like phase but decreases marginallyin the water-like phase. Table 3 also lists theτ0 values of thedyes in pure heptane, ethanol, and water, where the decays showclearly single-exponential behavior. It is seen that theτ0 valuesof the dyes in the water-like phase of the reverse micelles arequite close to theτ0 values in ethanol solution as expected fordyes localized at the surfactant head group region. Withincreasingw0, as the water pool size increases, the hydrationof the surfactant head groups also increases, causing a smallreduction in theτ0 values of the dyes in the water-like phase Itis interestingly seen that theτ0 values of the dyes in the heptane-like phase are significantly lower than those in pure heptane.This can be explained by considering that in the heptane-likephase the dyes are predominantly located in the nonpolar shellsof the reverse micelles. As mentioned earlier, it is likely thatthis shell contains a small amount of dispersed water molecules.The interaction of these water molecules with the dyes can causetheτ0 values in the heptane-like phase to be lower than that inpure heptane. On increasingw0, along with an increase in thewater pool size, there is an increase in the aggregation numberof the reverse micelles, and consequently the number ofhydrocarbon chains present in the nonpolar shell also increases.This possibly leads to a reduction in the number of dispersedwater molecules in this shell, causing theτ0 values of the dyes

Figure 5. Time-resolved fluorescence decays of C522 in AOT reversemicellar solution (w0 ) 20) measured at (a) 440 nm and (b) 510 nm inthe presence of (1) 0.0, (2) 0.006, (3) 0.02, and (4) 0.06 dm3 mol-1

DMAN. L is the lamp profile.

τav ) A1τ1 + A2τ2 (10)


to increase on increasingw0. It should be mentioned that theτ0

values of the dyes in pure heptane solution do not show anyobservable reduction on addition of AOT alone; the reductionbecomes pronounced only for the AOT/heptane/water ternarysystem. Thus lowerτ0 values for the heptane-like phase certainlyindicate that the dyes preferentially reside in the nonpolar shellof the reverse micelles rather than in bulk heptane.

As stated previously, theτav values of the dyes in both theheptane-like and the water-like phases gradually decrease asthe amine concentration is increased in the solution (cf.Figure 5). To estimate the quenching kinetics from TR measure-ments, theτav values were correlated using the SV relation asgiven by eq 11

The SV plots in the water-like phase for different coumarin-AN and coumarin-DMAN systems at a particularw0 valuesare shown in Figure 6. Unlike the SS fluorescence quenchingresults, the SV plots from TR measurements are seen to followthe proposed linearity for most of the coumarin-amine systems.As expected, in the TR measurements none of the SV plotsshowed any positive deviation. However, for some of thecoumarin-DMAN systems in the water-like phase, the SV plotsshowed negative deviations (cf. Figure 6b). This is possibly dueto the saturation effect of the DMAN concentration in the water-like phase, as discussed earlier in relation to the similar negativedeviations in SV plots from SS quenching. Thus, to avoid anycomplication, for those coumarin-amine systems that showednegative deviations in the SV plots, we used the initial slopesof the plots to estimate thekq values. For an actualkq

determination it is necessary to know the exact local concentra-tion of the quenchers, but the estimation of this parameter wasbeyond the scope of the present study. So we could only estimatethe relativekq values for different coumarin-amine systems,using the experimental amine concentrations directly. Thekq

values thus estimated for different coumarin-amine pairs arelisted in Table 3. Though thesekq values are not absolute, theycan be used safely for comparisons as well as for the correlationwith the energetics of the ET reactions (Marcus correlation) byconsidering each of the amines separately. Marcus correlationsof the present systems will be discussed in the later part of thissection. At present, it is interesting to compare thekq values inAOT reverse micelles with those reported for similar coumarin-amine systems in a homogeneous solution under diffusiveconditions.44

Although the anisotropy decay of the dyes in the water-likephase of the reverse micelles is quite complex in comparisonto that in a homogeneous acetonitrile solution, yet from theconsideration that the average rotational times in the water-like phase (e.g., 851 ps for C153 atw0 ) 40) is about 30 timeshigher than theτrot of the dyes in acetonitrile solution(∼30 ps),38 we can expect that the microviscosity around thesurfactant headgroup region of the micelle could also be morethan 1 order of magnitude higher than that of acetonitrile.Considering that the reported bimolecular diffusion-controlledrate constant,kd, for coumarin-amine systems in acetonitrilesolution is∼1.5× 1010 dm3 mol-1 s-1, thekd in the water-likephase of the reverse micelles can at most be expected to be onthe order of 5× 108 dm3 mol-1 s-1. In fact, in the water-likephase the approach of the quenchers to the fluorophore will

TABLE 3: Lifetimes of the Coumarin Derivatives in Pure Heptane, Water, Ethanol, and the Two Phases of AOT ReverseMicelles as Well as the Corresponding Bimolecular Quenching Constants for DMAN and ANa

τ0

(ns)τ0

(ns)kq (DMAN)/109

(dm3 mol-1 s-1)kq (AN)/109

(dm3 mol-1 s-1)

coumarin heptane water EtOH w0 heptane-like water-like heptane-like water-like heptane-like water-like

C153 3.86 2.56 4.70 10 2.65 4.12 12.66 1.70 11.86 4.2120 2.78 3.99 15.04 1.58 11.8 4.4240 2.88 3.86 15.07 1.53 11.95 4.63

C522 3.70 3.25 5.10 10 2.53 4.51 21.85 2.18 13.39 5.2120 2.61 4.42 21.90 2.06 12.68 5.2940 2.73 4.36 23.85 1.84 12.85 5.65

C152 3.50 0.46 1.63 10 2.01 2.86 27.26 1.97 13.85 6.3120 2.26 2.88 26.47 1.94 14.85 6.9240 2.53 2.89 27.45 1.79 13.96 6.91

C151 0.90 4.56 5.3 10 5.36 1.58 4.5720 5.23 1.45 4.8940 5.19 1.40 4.87

a The values are correct within(5% experimental error.

τ0

τav) 1 + kqτ0[Q] (11)

Figure 6. Stern-Volmer plots for time-resolved fluorescence quench-ing of C151 (9), C152 (b), C153 (0), and C522 (O) by (a) AN and(b) DMAN in the water-like phase of AOT reverse micelles atw0 ) 20.


not be equally probable from all directions as otherwise possiblein a homogeneous solution. Accordingly, the effectivekd in thepresent cases can be even lower than 5× 108 dm3 mol-1 s-1.From the observation that thekq values for some coumarin-amine systems in the water-like phase are>5 × 109 dm3 mol-1

s-1, it is indicated that the diffusion of the reactants does notplay any significant role in determining the quenching kineticsin the present systems. We feel that as in normal micelles theinteraction in the coumarin-amine systems in the water-likephase of the reverse micelle is also nondiffusive in nature andthe quenching kinetics are mainly determined by the spatialdistribution of the amine quenchers around the excitedfluorophore.25-27,45,46Under this condition, as there will be adistribution in the donor-acceptor separations, the observedfluorescence decays of the dyes in the presence of quencherwill be inherently non-single-exponential in nature.25-27,45,46Inthe present systems, however, due to the complexity of thereverse micelle structure and its inherent inhomogenity even ina particular phase, a detailed analysis of the fluorescence decayson the basis of the quencher distributions is extremely difficult.Thus, to avoid complexity, we adopted a simpler approach toanalyze the fluorescence decays as a biexponential function,and the average lifetimes are used to determine the effectivequenching constantkq following eq 11.

An inspection of thekq values in Table 3 reveals a numberof interesting details. It is seen that thekq values in the water-like phase are much lower than those in the heptane-like phasefor both of the quenchers. This may be understood by consider-ing that the dyes present in the nonpolar shell of the reversemicelles can be quenched not only by the amines present inthis region but also by those present in the bulk heptane outsidethe reverse micelles. On the contrary, the dyes present in thesurfactant head group region of the water-like phase can bequenched only by the amines solubilized in the same region ofthe micelle. Accordingly,kq in the heptane-like phase is muchhigher than that in the water-like phase. In the heptane-likephase, it is observed that thekq values follow the order DMAN> AN. This is in accordance with the oxidation potentials ofthe amines concerned. Considering that theτrot in the heptane-like phase is about 6-7 times higher than that in acetonitrilesolution, the expectedkd value in the former cases should havebeen around 2× 109 dm3 mol-1 s-1. It is observed that thekq

values in the heptane-like phase are much higher than this value.This is possibly because in this region also the quenchingkinetics are largely determined by the spatial distribution of theamine quenchers around the excited fluorophore rather than bythe mutual diffusion of the reactants. Contrary to the heptane-like phase, thekq values in the water-like phase are seen tofollow the order AN> DMAN. This reversal in the trend forthe kq values between AN and DMAN can be rationalized onthe basis of the relative solubilities of the two quenchers in thewater-like and heptane-like phases and their water/heptanepartition coefficients. As DMAN is much more hydrophobic innature, its solubility in the water-like phase is much less thanthat of AN. This reduced solubility is the reason for the lowerkq values for DMAN even though the oxidation potential ofDMAN is more favorable than that of AN for the ET reactionwith the coumarin dyes.

Another interesting observation from the present results isthat in the water-like phase, with increasingw0, the kq valuesfor DMAN show a marginal decrease but thekq values for ANshow a slight increase. Since the partition coefficient for DMANin water is very low, the reduction inkq with increasingw0 canbe explained in terms of the reduced solubility of the quencher

in the water-like phase. However, as AN has a higher affinityfor the water-like phase, its solubility possibly increases to someextent on increasingw0, resulting in a marginal increase in thekq values. That AN has a good affinity for the water-like phaseis also indicated by the fact that the SV plots from SSmeasurements in this phase always show a positive deviationfrom linearity on using AN as the quencher.

Finally, it is interesting to see how the experimentalkq valuescorrelate with the free energy changes (∆G°) for the ETreactions. The∆G° values for the present coumarin-aminesystems were calculated according to the Rehm-Weller relationas47

whereE(D/D+) andE(A/A-) are the oxidation and reductionpotentials of the amines and coumarins, respectively,E00 is theexcitation energy of the coumarin dyes in the S1 state,e is thecharge of an electron, andr is the center-to-center contactdistance between the coumarin and the amine. The radii of themolecules were estimated using Edward’s volume additionmethod.48 Theε value for the water-like phase was approximatedas the dielectric constant of ethanol since theτ0 values of thedyes in the water-like phase of the micelle were quite close tothe values in ethanol solution. Similarly, theE00 values of thedyes in the water-like phase were estimated from the overlapof the excitation and emission spectra of the dyes in ethanolsolution. For the heptane-like phase, theε andE00 values wereconsidered using a heptane solution of the dyes as the reference.The redox potentials of the coumarin acceptors and aminedonors in the water-like phase were estimated following theprocedure given in section 2. Since the redox potentials of theacceptors and donors could not be estimated in heptane solution,we used the reported values in acetonitrile solution directly soas to have a relative estimate of the∆G° values and to see atleast qualitatively the nature of the Marcus correlation curvesin the heptane-like phase. All of the ET-related parameters forthe present systems are listed in Table 4.

The kq versus∆G° correlations for the coumarin-DMANand coumarin-AN systems for the water-like phase at differentw0 values are shown in Figure 7. As predicted by Marcus ETtheory, clear inversions in thekq values are observed for thesesystems at the higher exergonicity region. The inversionsapparently occur at∆G° ≈ 1.2-1.3 eV, which is the expectedexergonicity region for inversion considering that the solventreorganization energy,λs, for ET reactions in a polar mediumis ∼1 eV.27,28This also indicates that in the water-like phase ofthe reverse micellesλs contributes fully toward the activationbarrier for the ET reactions.25,26Such a situation is expected ifthe solvation dynamics in the water-like phase are much fasterthan the observedkq values.25 Riter et al.16 have indeed observedsolvation times of∼0.16 and∼2.2 ps in reverse micelles, whichaccounts for the majority of the Stokes shifts due to solventrelaxation. Considering these solvation times, the solventrelaxation rates in the water-like phase of the reverse micellecan be as high as∼5 × 1011 s-1, much higher than the observedkq values.

Recently some results on ET reactions in reverse micelleshave been reported where the inversion observed in the ET ratesat higher exergonicity has been attributed to the heterogeneousdistribution of the probes and the consequent differences in theirdiffusion rates.49 In the present work, we have ascertained thatin the water-like phase the coumarin dyes are solubilized mainlyin the interfacial region in strong association with the surfactant

∆G° ) E(D/D+) - E(A/A-) - E00 - e2

εr(12)


head groups where the diffusion of the reactants is much slowerthan the observed ET rates. Moreover, theτw andτslow valuesof the anisotropy decays as estimated for the dyes in the water-like phase (cf. Table 3) do not show any inversion-like behaviorwith the exergonicity of the ET reactions. A comparison of theanisotropy results and the ET results involving C153 and C522dyes reveals that although the rotational correlation times forC153 are lower thekq values are actually higher for the C522dye, a result just opposite to that expected on the basis of thediffusion rates of the dyes in the micellar phase. So highermobility does not necessarily ensure a higherkq value in thepresent systems. We strongly feel that the ET interaction in thecoumarin-amine systems in reverse micelles is nondiffusivein nature and the quenching kinetics in the present systems aremainly determined by the spatial distribution of the quencher

molecules around the excited fluorophores, as has been observedby us earlier in normal micelles25-27 and is also discussed byMorandeira et al. for ET reactions in electron donatingsolvents.45,46 A similar aspect has also been suggested byTavernier et al. for quenching kinetics in micellar media.50 Ifthe retardation in the ET rates at higher exergonicity had beendue to the differences in reactant diffusion or due to thedifferences in the accessibility of the quenchers for differentfluorophores, thenkq values should have been randomlydistributed with ∆G° rather than showing a clear inversionbehavior. Thus, it is evident that the observed inversion in thekq versus∆G° correlations for the ET reactions in reversemicelles is a real Marcus inversion and arises due to theenergetics of the reactions as envisaged from Marcus ET theory.

As mentioned earlier, one of the reasons for not observingthe Marcus inversion behavior for intermolecular ET reactionsis the lack of availability of suitable donor-acceptor pairs thatcan attain sufficiently high exergonicity. In the present study,the coumarin-AN and coumarin-DMAN systems in the water-like phase appear to be good systems in achieving reasonablyhigh exergonicity for the observation of the inversion behavior.The nondiffusive nature of the ET reactions in the micellarmedia further helps in observing the inversion behavior as theET in these systems effectively occurs as equivalent to unimo-lecular reactions with negligible influence of the reactantdiffusion. That diffusion of the reactants does not play asignificant role is also attested by the very marginal variationsin thekq values withw0. The Marcus correlation curves also donot show any marked variation withw0. The correlation curvesare raised marginally in the case of the coumarin-AN systemsand lowered marginally in the case of the coumarin-DMANsystems on increasingw0. These changes are possibly due to

TABLE 4: Redox Potentials, E00, and ∆G0 Values in Heptane-like and Water-like Phases

E(C/C-) vs SCE(V)

E00

(eV)E(A/A -) vs SCE

(V)∆G°(eV)

coumarin heptane-likea water-like heptane-like water-like amines heptane-likea water-like heptane-like water-like

C153 -1.69 -1.14 2.92 2.48 AN 0.93 0.39 -0.33 -0.97DMAN 0.72 0.22 -0.54 -1.14

C522 -1.659 -1.11 3.00 2.55 -0.44 -1.08-0.65 -1.25

C152 -1.63 -1.08 3.08 2.61 -0.55 -1.17-0.76 -1.33

C151 -1.02 2.82 -1.43-1.60

a Due to the unavailability of a suitable supporting electrolyte for heptane, redox potentials in acetonitrile were used directly for the heptane-likephase for a qualitative correlation.

Figure 7. The ln(kq) vs ∆G° plots for (a) coumarin-AN and (b)coumarin-DMAN systems atw0 ) 10 (b) and 40 (O) in AOT reversemicelles monitored for the water-like phase.

Figure 8. The ln(kq) vs ∆G° plots for coumarin-AN (s) andcoumarin-DMAN (---) in AOT reverse micelles monitored for theheptane-like phase.


the changes in solubility of AN and DMAN in the water-likephase of the reverse micelles with increasing water pool size.Since the coumarin dyes in the water-like phase are preferen-tially solubilized in the interfacial region, the effect ofw0 isnot reflected appreciably on thekq values. If the dyes weresolubilized deep inside the water pool, much higher variationin the kq values could be expected on changingw0.

For the heptane-like phase, only the rising part (normalregion) of thekq versus∆G° correlation curves was observed(cf. Figure 8) especially because the data point for the invertedregion at high exergonicity could not be obtained due tononsolubility of the C151 dye in this phase. As discussed earlier,the kq values in the heptane-like phase are also much higherthan expected for a diffusion-controlled reaction. So it isproposed that in this phase too ET kinetics are primarilygoverned by the spatial distribution of the quenchers aroundthe excited fluorophores rather than by the diffusion of thereactants. In this situation, it is expected that the inversionbehavior could also have been observed in this phase if suitabledonor-acceptor pairs were available for extending the studytoward higher exergonicity.

Conclusion

Reverse micelles provide us the opportunity to explore ETkinetics in the different regions where the reactants can localize,namely, the heptane-like and water-like phases. Steady-statestudies, time-resolved fluorescence, and anisotropy results revealthat the coumarin dyes in the water-like phase are actuallylocated in the surfactant-water interfacial region and in theheptane-like phase they are mainly present in the nonpolar shellcomposed of the surfactant tails and intervening heptanemolecules. The amine quenchers are also distributed betweenthe two phases depending on their hydrophobicities and water/heptane partition coefficients. It is seen that the water-like phasepresents a suitable environment for the observation of theMarcus inverted region for bimolecular ET reactions by arrestingthe diffusion of the reactants. Because of efficient solvation inthis region, the solvent reorganization energy contributescompletely to the activation barrier for the ET reactions, andthe inversion occurs at a high exergonicity as expected in apolar media. Due to the location of the dyes in the interfacialregion, variation ofw0 values of the reverse micelles does nothave any profound influence on the Marcus correlation curves.The general nature of the curves and the inversion region arefound to remain similar at allw0 values. The small effect ofw0

on the quenching constants in the water-like phase is found tobe partly dependent on the nature of the amines and theirpartition coefficients. In the heptane-like phase too, diffusiondoes not play any major role, and the ET kinetics are primarilydetermined by the spatial distribution of the reactants. So Marcusinversion behavior could have been observed in this phase alsoif the studies could be further extended toward the higherexergonicity region. Finally, it is to be emphasized that thougha substantial amount of photochemical work has been carriedout in reverse micelles the reports on ET reactions in theserestricted media, especially in relation to the realization of theMarcus inversion behavior for bimolecular ET reactions, arevery limited. It is evident from the present results that thecompartmentalization of the reactants in reverse micellesprovides a conducive environment for the Marcus inversionbehavior to be observed easily for bimolecular ET reactions.These results in AOT reverse micelles, thus, strengthen ourearlier inference25-27 that the bimolecular ET reactions inrestricted environments effectively occur under nondiffusive

conditions, and this leads to the easy observation of the Marcusinversion behavior for such reactions, though such an inversionbehavior is normally obscured for bimolecular ET reactions inhomogeneous solution.

Acknowledgment. S. D. C. acknowledges the Departmentof Atomic Energy and Board of Research in Nuclear Sciencesfor the Dr. K. S. Krishnan Research Associateship. The authorsare thankful to Mr. A. K. Satpati of the Analytical ChemistryDivision, Bhabha Atomic Research Centre, for his kind help inthe cyclic voltammetric measurements.

References and Notes

(1) ReVerse Micelles; Luisi, P. L., Straub, B., Eds.; Plenum Press: NewYork, 1984.

(2) Structure and ReactiVity in ReVerse Micelles; Pileni, M. P., Ed.;Sudies in Physical and Theoretical Chemistry 65; Elsevier: Amsterdam,1989.

(3) Luisi, P. L.; Giomini, M.; Pileni, M. P.; Robinson, B. H.Biochim.Biophys. Acta1988, 947, 209.

(4) Wilcoxon, J. P.; Williamson, R. L.; Baughman, R.J. Chem. Phys.1993, 98, 9933.

(5) Silber, J. J.; Biasutti, A.; Abuin, E.; Lissi, E.AdV. Colloid InterfaceSci.1999, 82, 189.

(6) Belletete, M.; Lachapelle, M.; Durocher, G.J. Phys. Chem.1990,94, 5337.

(7) Zinsli, P. E.J. Phys. Chem.1979, 83, 3223.(8) Hasegawa, M.; Sugimura, T.; Suzaki, Y.; Shindo, Y.; Kitahara, A.

J. Phys. Chem.1994, 98, 2120.(9) Jain, T. K.; Varshney, M.; Maitra, A.J. Phys. Chem.1989, 93,

7409.(10) Onori, G.; Santucci, A.J. Phys. Chem.1993, 97, 5430.(11) Christopher, D. J.; Yarwood, J.; Belton, P. S.; Hills, B. P.J. Colloid

Interface Sci.1992, 152, 465.(12) Hauser, H.; Haering, G.; Pande, A.; Luisi, P. L.J. Phys. Chem.

1989, 93, 7869.(13) Zhang, J.; Bright, F. V.J. Phys. Chem.1991, 95, 7900.(14) Cho, C. H.; Chung, M.; Lee, J.; Nguyen, T.; Singh, S.; Vedamuthu,

M.; Yao, S. H.; Zhu, J. B.; Robinson, G. W.J. Phys. Chem.1995, 99,7806.

(15) Sarkar, N.; Das, K.; Datta, A.; Das, S.; Bhattacharyya, K.J. Phys.Chem.1996, 100, 10523.

(16) Riter, R. E.; Willard, D. M.; Levinger, N. E.J. Phys. Chem. B1998, 102, 2705.

(17) Yamasaki, T.; Kajimoto, O.; Hara, K.J. Photochem. Photobiol., A2003, 156, 145.

(18) Kalyanasundaram, K.Photochemistry in Microheterogeneous Sys-tems; Academic Press: London, 1987.

(19) Grand, D.J. Phys. Chem. B1998, 102, 4322.(20) Grand, D.; Dokutchaev, A.J. Phys. Chem. B1997, 101, 3181.(21) Borsarelli, C. D.; Cosa, J. J.; Previtali, C. M.Photochem. Photobiol.

1998, 68, 438.(22) Borsarelli, C. D.; Bertolotti, S. G.; Previtali, C. M.Photochem.

Photobiol.2001, 73, 97.(23) Aoudia, M.; Rodgers, M. A. J.J. Phys. Chem. B2003, 107,

6194.(24) Borsarelli, C. D.; Cosa, J. J.; Previtali, C. M.Langmuir1993, 9,

2895.(25) Kumbhakar, M.; Nath, S.; Mukherjee, T.; Pal, H.J. Chem. Phys.

2005, 123, 034705.(26) Kumbhakar, M.; Nath, S.; Pal, H.; Sapre, A. V.; Mukherjee, T.J.

Chem. Phys.2003, 119, 388.(27) Kumbhakar, M.; Mukherjee, T.; Pal, H.Chem. Phys. Lett.2005,

410, 94.(28) Kavarnos, G. J.Fundamentals of Photoinduced Electron Transfer;

VCH Publishers: New York, 1993.(29) Electron Transfer in Inorganic, Organic, and Biological Systems;

Bolton, J. R., Mataga, N., McLendon G. L., Eds.; Advances in ChemistrySeries 228; American Chemical Society: Washington, DC, 1991.

(30) Miller, J. R.; Calcaterra, L. T.; Closs, G. L.J. Am. Chem. Soc.1984, 106, 3047.

(31) Miller, J. R.; Beitz, J. V.; Huddleston, R. K.J. Am. Chem. Soc.1984, 106, 5057.

(32) Tayar, N. E.; Testa, B.; Carrupt, P.-A.J. Phys. Chem.1992, 96,1455.

(33) Lakowicz, J. R.Principles of Fluorescence Spectroscopy, 2nd ed.;Kluwer Academic/Plenum Publishers: New York, 1999.


(34) The reportedE(D/D+) andE(A/A-) values in acetonitrile solutionversus a saturated calomel electrode (SCE) were reduced by 0.542 V toobtain the corresponding values in an AOT reverse micellar solution.Thisshift in the electrode potentials (i.e., 0.542 V) was estimated from theaverage shift in theE(A/A-) potential of the coumarin dye, C151, which ismostly localized in the water-like phase in AOT, compared to that inacetonitrile.35-37

(35) Pal, H.; Shirota, H.; Tominaga, K.; Yoshihara, K.J. Chem. Phys.1999, 110, 11454.

(36) Jones, G., II.; Griffin, S. F.; Choi, C. Y.; Bergmark, W. R.J. Org.Chem.1984, 49, 2705.

(37) Castner, E. W., Jr.; Kennedy, D.; Cave, R. J.J. Phys. Chem. A2000, 104, 2869.

(38) The anisotropy decays for the C153 and C522 dyes in acetonitrilehave been measured by a fluorescence up-conversion setup with aninstrument response function of∼230 ps at fwhm.

(39) Quitevis, E. L.; Marcus, A. H.; Fayer, M. D.J. Phys. Chem.1993,97, 5762.

(40) Maiti, N. C.; Krishna, M. M. G.; Britto, P. J.; Periasamy, N.J.Phys. Chem. B1997, 101, 11051.

(41) Dutt, G. B.J. Phys. Chem. B2004, 108, 805.(42) Wittouck, N.; Negri, R. M.; Ameloot, M.; De Schryver, F. C.J.

Am. Chem. Soc.1994, 116, 10601.(43) Andrade, S. M.; Costa, S. M. B.Photochem. Photobiol. Sci.2003,

2, 605.(44) Nad, S.; Pal, H.J. Phys. Chem. A2000, 104, 673.(45) Morandeira, A.; Fu¨rstenberg, A.; Gumy, J. C.; Vauthey, E.J. Phys.

Chem. A2003, 107, 5375.(46) Morandeira, A.; Fu¨rstenberg, A.; Vauthey, E.J. Phys. Chem. A

2004, 108, 8190.(47) Rehm, D.; Weller, A.Isr. J. Chem.1970, 8, 259.(48) Edward, J. T.J. Chem. Educ.1970, 47, 261.(49) Chakraborty, A.; Seth, D.; Chakrabarty, D.; Hazra, P.; Sarkar, N.

Chem. Phys. Lett.2005, 405, 18.(50) Tavernier, H. L.; Laine, F.; Fayer, M. D.J. Phys. Chem. A2001,

105, 8944.


Documents

Compartmentalization of Reactants in Different Regions of Sodium 1,4-Bis(2-ethylhexyl)sulfosuccinate/Heptane/Water Reverse Micelles and Its Influence on Bimolecular Electron-Transfer