Nanosecond time resolved emission spectroscopy of aminocoumarins in AOT reversed micelles B. Bangar Raju§ and Silvia M. B. Costa* Centro de Quimica Estrutural, Complexo I, Instituto Superior T ecnico, Av. Rovisco Paris 1, 1049-001, L isboa Codex, Portugal. E-mail : FAX:]351-1-846 44 55 sbcosta=alfa.ist.utl.pt, Received 30th July 1999, Accepted 10th September 1999 Time resolved emission spectroscopy of a water insoluble aminocoumarin derivative, BC I, in n-heptane»AOT»water reversed micelles is investigated using a nanosecond —uorescence spectrometer. The extent of the time-dependent Stokes shift is found to be smaller for than for and also for W0\4 W0\10 molar ratio of water and AOT).A monoexponential decay of the dynamic Stokes shift W0\40 (W0 , correlation function with a time constant of 4.5 ns is observed at At the time-correlation W0\4. W0\10 function shows a bi-exponential decay with a fast component of D600 ps and another of 3.3 ns comparable to the —uorescence lifetime of the dye.The time correlation function at shows an extremely slow W0\40 component of the solvation time which is at least an order of magnitude larger than the —uorescence lifetime of the probe dye. In addition a faster component with a solvation time of 2.15 ns is also observed. The solvation time observed is likely due to the diÜusion of the probe dye within the micellar interface or due to the reorganization of the water molecules bound to the polar head groups of AOT.Steady-state absorption and —uorescence, and time resolved —uorescence study of another aminocoumarin, C 480, give evidence of equilibrium between the distribution of the dye molecules between the interface and water pool. The time resolved shift of the emission spectra is more likely due to the diÜusion of the probe between the interface and water pool rather than reorientation of the water molecules (solvent relaxation) in the core of the reversed micelles as it has been reported earlier.Introduction Water is a major and an important constituent of biological membranes and determines the dynamics of biological processes. Solvation dynamics in restricted media is now being widely investigated in an eÜort to understand the ultrafast processes that take place in biological membranes. This was stimulated by the depth of information obtained about the relaxation times of pure solvents and solvent mixtures using picosecond (ps) and femtosecond (fs) laser spectroscopy tools.1 While the solvent relaxation time of bulk water is in the subpicoseconds, in microheterogeneous media it is shown to vary, depending on the nature of water (free or trapped/bound), between few picoseconds and nanoseconds (ns).2h8 Aminocoumarin and anilinonaphthalene sulfonate (ANS) dyes are some of the most widely used probes to study solvation dynamics in restricted heterogeneous media, such as cyclodextrins,2 liquid crystals,3 micelles,4 vesicles,5 zeolites6 and reversed micelles.7h9 Location of the probe and the surroundings it experiences constitute an important aspect in —uorescence probe studies, whether steady-state or timeresolved, of microheterogeneous systems.For instance in Aerosol OT, AOT [sodium 1,4-bis(2-ethylhexyl) sulfosuccinate], reversed micelles the solubility of the dye and its preferential solvation, if any, at a speci–c site determine its absorption and emission characteristics.When a dye is soluble in all the three pseudophases of the reversed micelles, the distribution and diÜusion of the dye between the three diÜerent microenvironments and the partition of the dye as well as the § Present address : Department of Chemistry, University of Connecticut, 55 North Eagleville Road, Storrs, Connecticut 06269-3060, USA. diÜusion of the water molecules, need to be taken into consideration. This is in addition to the overall motion of the reversed micelles and the wobbling and tumbling motion of the probe dye in the reversed micelles.10 The aminocoumarin and ANS dyes that have so far been used to probe solvation dynamics in n-heptane»AOT»water reversed micelles are soluble in water as well as in alkane» AOT binary solutions.Hence the study of solvation dynamics is complicated due to the diÜusion of water molecules and, sometimes, also of the probe dye, between the oil, interface and the water pool. It is essential, hence, to characterize the location of the probe and whether or not there is a partitioning of the dye in the reversed micelles before extending the investigations to determine the time evolution of the —uorescence spectrum.We have attempted to investigate solvation dynamics of a water insoluble aminocoumarin dye, BC I, (Scheme 1) in nheptane »AOT»water reversed micelles. BC I is a non-rigid aminocoumarin probe dye with a —exible diethylamino and a cyano group as substituents at the 7- and 4-positions, respectively, of the parent coumarin.A second coumarin moiety is also present as a substituent at the 3-position. From the measurements of —uorescence excitation spectra at diÜerent emission wavelengths, steady-state —uorescence emission anisotropy and —uorescence decays at diÜerent excitation and emission wavelengths, it was recently shown by us11 that the dye experiences two diÜerent environments within the micellar interface. The dye is strongly in—uenced»due to hydrogen bonding»by the water molecules bound to the polar head groups of the surfactant.It was also shown that it is possible to selectively excite only those dye molecules that are located in the proximity of the polar head groups.11 Phys. Chem. Chem. Phys., 1999, 1, 5029»5034 5029 This journal is The Owner Societies 1999 (Scheme 1 Structures of the two 7-diethylaminocoumarin dyes discussed in the presented study. In this paper we discuss the time resolved —uorescence (nanosecond resolution) of BC I in the interfacial region of AOT reversed micelles for diÜerent values of W0 (W0\ We also revisit the reported solvation [water]/[AOT].dynamics of a water-soluble rigid aminocoumarin dye, C 480,7 (Scheme 1). A critical examination of the —uorescence spectrum of C 480 in AOT reversed micelles given as Fig. 2 in ref. 7 shows an isoemissive point in the region where there is no contribution of the —uorescence of the dye molecules in nheptane.This then implies that there is an equilibrium between the distribution of the dye molecules in the interfacial region and water pool of the reversed micelles. We hope to show that the results thus obtained can be better explained by taking into account the partition of this probe between the water and the interface rather than solvent relaxation, as reported earlier.7 Experimental The synthesis of BC I (a cyano derivative of coumarinyl benzopyrano pyridine) has been reported earlier.12 This was later puri–ed using column chromatography (silica gel, dichloromethane»ethylacetate).The purity of the dye has been veri–ed by thin layer chromatography on silica gel with the above solvent mixture as carrier. AOT is ultra pure grade from Sigma and n-heptane of spectroscopy grade is from Aldrich. Coumarin 480 is from Exciton. The three chemicals have been used as received. Water used was doubly distilled. The mole fraction of intrinsic water to AOT as deter- (W0), mined by Karl-Fischer titration, was found to be 0.2.Absorption spectra were recorded using Jasco V 500 UV/VIS/NIR absorption spectrometer. Steady-state —uorescence spectra were measured on Perkin Elmer 3B spectrometer with the cuvette holder placed in a perpendicular geometry. The spectral correction –le supplied by the manufacturer was used to obtain the corrected —uorescence spectrum. All the measurements were done at room temperature. Fluorescence lifetimes were determined with a PTI LS-1 (qf) time correlated single photon counting instrument using a nanosecond hydrogen —ash lamp.13 Fluctuations in the pulse jitter and intensity were corrected by making an alternate collection of scattering and sample emissions. In this way 10% of the total number of counts at the maximum were collected each time.Data analysis was carried out using the curve- –tting program supplied by the manufacturer. Reduced chisquared, s2, (0.95\s2\1.1) and a high Durbin»Watson parameter ([1.7) were considered while determining the quality of the –t.13 The error in the determination of the —uorescence lifetimes is ^200 ps.A utility in the software used to control the instrument is applied to generate —uorescence decays at diÜerent wavelengths and also to determine the maximum intensity of the decay at each measured wavelength. Spectrum reconstruction is performed following the method proposed by Maroncelli and Fleming.14 Brie—y, —uorescence decays are collected between 560 and 650 nm and in intervals of 10 nm.The data are then –t to a sum of exponentials. Then using the pre-exponential factors and the lifetimes, deconvoluted decays at the measured wavelengths are reconstructed. As the time-correlated single photon counting data is collected at a constant signal-to-noise ratio, the calculation of the integrated emission intensity is accomplished by scaling to the data acquisition time at each wavelength since this time is directly proportional to the integrated emission intensity.î Peak frequencies are determined by –tting the spectral points at the times of interest to a log-normal line shape function.14 The —uorescence maximum found in the steady-state spectrum is taken as the peak frequency at in–nite time.Results and discussion (a) Emission spectra of BC I at the AOT interface The spectroscopic characteristics of BC I have recently been investigated mainly in neat aprotic solvents15 and in dioxane» water mixtures.16 These have been explained using the intramolecular charge transfer (ICT) and the twisted ICT (TICT) hypothesis.Studies in dioxane»water mixtures also give evidence of a weak intermolecular hydrogen bond between the cyano nitrogen and water in the ground state. The dye is also found to show considerable —uorescence quenching with increase in the concentration of water in the solvent mixture. The non-radiative deactivation rates have been correlated with the rate of intramolecular charge transfer processes in the dye.Austin Model 1 (AM 1) semiempirical calculations recently done on another dye belonging to the same family show that in such dyes the ground state is nearly isoenergetic (S0) whether the second coumarin moiety is in-plane or twisted out-of-plane with the parent coumarin.17 The energy of the lowest excited state is however found to be lowest when (S1) the bulky substitution containing a second coumarin is coplanar to the parent coumarin containing the diethylamine and the cyano groups thus extending the p-electron conjugation in the excited singlet state.17 Such a probability, in addition to the dependence of the photophysics of BC I on the concentration of water in dioxane,16 thus, make an interesting study of the intramolecular charge transfer processes of the dye in con- –ned systems, such as reversed micelles.11 The location of the probe dye, BC I, and its photophysical properties in reversed micelles were investigated recently using steady-state and time dependent —uorescence techniques.The –ndings are summarized as follows :11 BC I, a hydrophobic aminocoumarin derivative, is preferentially solvated in the organic phase of the reversed micelle. However, in view of its size it experiences a multitude of microheterogeneous environments. In the reversed micelles, emission arising from two diÜerent environments for the dye is observed.In n-heptane BC I shows a partly structured absorption spectrum in the visible region. Addition of 0.1 M of AOT to n-heptane solution of BC I results, among other changes, in a signi–cant broadening of the absorption spectrum at the rededge (j[520 nm) region where the dye shows practically no absorption in n-heptane solutions. On addition of water to the reversed micelles further broadening, albeit small, of the absorption spectra at the red-edge and a decrease in the absorbance around 470 nm was observed.Selective excitation î As described in LS-100TM —uorescence lifetime analysis PTI} module reference manual 5030 Phys. Chem. Chem. Phys., 1999, 1, 5029»5034of the dye molecules absorbing at a wavelength greater than 500 nm resulted in a single broad —uorescence emission with a peak at about 572 nm for The —uorescence efficiency W0\0. was found to decrease with increase in A bathochromic W0 . shift of about 25 nm was also observed on increasing No W0 .—uorescence was observed on excitation of BC I around this wavelength ([500 nm) in normal n-heptane solutions. This implies that the —uorescence spectrum hence obtained is contributed by BC I molecules that are solubilized in the interfacial region of the reversed micelle and that absorb at wavelengths greater than 520 nm. Fluorescence decays obtained for excitation at 540 nm were all mono-exponential, varying from 4.2 ns at to 3.0 nm at The W0\4 W0\40.steady-state —uorescence emission anisotropy of BC I, carried out at an excitation wavelength of 540 nm gives a non-zero value11 (mean value) which increases non-linearly with increase in These results give evidence that the dye mol- W0 . ecules solubilized in the interface of the reversed micelle contribute to the emission observed on excitation at 540 nm. The quenching of the —uorescence quantum efficiency and lifetime of the dye with increase in could be attributed to the W0 hydrogen bonding between the cyano group and the water molecules (bound water) in the micellar region of the interface. 11 (b) Time resolved emission spectra Time resolved —uorescence spectra of BC I in reversed micelles (at 10 and 40) are constructed by exciting the dye W0\4, at 540 nm. At time zero, peak wavelength observed is D580 nm, close to the maximum value in the steady-state spectrum for For the overall time resolved shift is only W0\2. W0\4 about 156 wavenumbers while it was found to be almost double (327 wavenumbers) for The analysis for W0\10.is complicated due to the presence of an extremely W0\40 slow component in the time correlation function. In order to make a quantitative analysis of the time resolved spectra in the reversed micelles the Stokes shift correlation function C(t) was calculated using the expression :14 C(t)\ l(t)[l(O) l(0)[l(O) (1) where, l(t), l(O) and l(0) are, respectively, the emission maxima at time t, in–nity (fully relaxed) and zero.The timeresponse functions of the dynamic Stokes shift observed at and 10 are shown in Fig. 1a. The best –ts obtained are W0\4 given as solid lines. At the mono-exponential –t of the W0\4, correlation function gives a time constant of 4.54 ns. At W0\ there is an initial fast component (around 600 ps) and a 10 much slower one of 3.3 ns. The latter value is close to the excited state lifetime of BC I at The corresponding W0\10. time evolution of the —uorescence spectra is shown in Fig. 1b. The time correlation function for does not relax to W0\40 zero value within the time frame of the experimental observation (Fig. 1c). It could however be –t with a biexponential function with solvation times (pre-exponential factors) of 2.15 ns (0.80) and 21.3 ns (0.2). We could not resolve any extremely short-lived component, even if it were to be present, in the solvent relaxation function at this value of A long-lived W0 . second component of the solvation time has also been observed, earlier by Sarkar et al.7 and by us in this study (see below) for a rigid aminocoumarin dye, C 480, in aqueous reversed micelles.More recently Shirota and Horie18 have also observed a similar slow solvent relaxation time in nonaqueous reversed micelles using a rigid aminocoumarin dye, C 343. This long-lived component is generally found to be much larger than the —uorescence lifetime of the dye. With increase in although the viscosity of the reversed micelle decreases, W0 , the dielectric constant and, more importantly, the rotational relaxation time of the micelle increases18 and becomes (qrM) Fig. 1 (a) Time-response function of the dynamic Stokes shift observed at and 10. (b) Time resolved emission spectra of BC W0\4 I in AOT reversed micelles at (c) Time correlation function W0\10. for W0\40. much longer than the rotational diÜusion of the probe dye in the reversed micelle. This could be one reason for the large oÜset observed in the C(t) function.This shows up as a dark or non-observable solvation dynamics. Shirota and Horie18 observed that the dark solvation dynamics component increases with increase in This is in agreement with the W0 . Phys. Chem. Chem. Phys., 1999, 1, 5029»5034 5031results obtained earlier by Zinsli19 using Smoluchowsky diÜusion equations. The reasons for the large oÜset leading to a longer second component in the time correlation function in certain microheterogeneous systems are, however, not well understood.Zhang and Bright9 have used a charge transfer probe (ANS) to investigate the nanosecond reorganization rates of water molecules using frequency-domain —uorescence spectroscopy. They found that the reorganization time of bound water molecules reduces from D16 ns at to 3.5 ns at W0\0 W0\2.8.20 The relaxation times in the reversed micelles at and 10 W0\4 obtained by us in the present study using BC I as the probe dye are close to those reported by Zhang and Bright for For free water in the water pool, the same authors W0\2.8.9 report9 that the relaxation time decreases from 1.08 ns at to D500 ps at These values are much faster W0\0 W0\8.3.than the 8 ns and 1.7 ns times obtained, respectively, at W0\ and by Sarkar et al.7 using a structurally rigid 4 W0\32 Coumarin C 480. The average relaxation time SqT“ obtained by Sarkar et al.7 at (6.75 ns) is almost the same as W0\32 that at (8 ns). This implies that the solvent relaxation W0\4 does not become faster with increase in This is contrary W0 .to what one observes in reversed micelles.8,9 Solvent relaxation times reported by Sarkar et al. could then be due to some processes other than solvation dynamics. This subject is discussed in section c. The faster decay of the time-correlation function at W0\40 and 10 than at obtained in the present study is consis- W0\4 tent with the reorganization times obtained by Zang and Bright,9 Vajda et al.2 using ICT probes.Recently, the latter authors2 have found that the solvation dynamics of Coumarin 460 (a TICT candidate-dye) in c-cyclodextrin is faster than that of the rigid aminocoumarin dye, C 480. Barbara and coworkers also observed that the dielectric relaxation times of the solvents was faster when observed using Coumarin 120 (7- dimethylamino-4-tri—uoromethyl coumarin) as the probe dye.21 In neat solvents, the solvation dynamics of C 120»a dye that is reported to undergo TICT in highly polar solvents21,22»has been reported to be 20 to 40% faster than the structurally rigid aminocoumarin dye, C 153.21 In reversed micelles ICT processes (correlated to the decrease in the —uorescence lifetime and increase in rate of non-radiative decay) are known to become faster with increase in W0 values.23 The polarity of the interior of the reversed micelles increases with increase in from 4 to 10.23,24 Increase in W0 polarity of the solvent medium is known to facilitate a greater charge transfer in BC I.15,16 The decrease in the —uorescence lifetime with increase in is due to the increase in W0 ICT.11,15,16 This speeds up the reorganization of the solvent around the newly formed solute dipoles in the excited state.The solvation time of BC I in reversed micelles at W0\40 and 10 is, however, three orders of magnitude slower than that observed in bulk water.8 As mentioned earlier, no isosbestic or isoemissive point is observed in the absorption and —uorescence spectra of BC I for diÜerent values of in AOT reversed micelles.In view of W0 its hydrophobic nature the dye resides solely in the oil and in the micellar interface. To our knowledge there has been only one earlier report of time-resolved —uorescence study of a dye present exclusively in the micellar interface. Pansu and coworkers25 studied the diÜusion kinetics of bianthryls (BA and BOA) in CTAC micelles in the excited state.Bianthryl, BA, is a TICT candidate dye26,27 and is a neutral molecule and insoluble in water. It is hence solubilized in the core of the hydrophobic domain of the micelle. The time-dependent Stokes shift of the —uorescence spectra showed two temporary isoemissive points for BA and one for BOA. The dynamic where A is the pre-exponential factor and q is “ SqT\A1q1]A2q2 , the lifetime. Stokes shift has been interpreted as due to a ììdiÜusion of the excited probe from the non-polar hydrophobic core to the water interface of the micelle where it is eventually trappedœœ.BC I is also a neutral molecule and is insoluble in water. In reversed micelles it shows a preferential solvation in the oil phase. Thus the relaxation times observed in the present study may either be due to the diÜusion of the dye molecules within the interface or to the reorganization of the highly constrained bound water molecules in the interface. (c) Study of C 480 in AOT reversed micelles We have also investigated the time resolved spectra of C 480 in AOT reversed micelles and obtained the time resolved emission spectrum after subtracting the dye molecules contribution in n-heptane.The relaxation times (1.7 ns and 12 ns) and the pre-exponential factors (0.5 and 0.5, respectively) obtained by us for C(t) using a nanosecond —uorescence spectrophotometer agree with those reported by Sarkar et al.7 for C 480 at Although the 1.7 ns component is within W0\32.the —uorescence lifetime of the dye molecule, the physical meaning of the dominant 12 ns component is not trivial to comprehend. It is more likely to be representative of a process associated with the translational or rotational reorientation or diÜusion time of the reversed micelle rather than solvation dynamics of the probe dye.19 In order to understand the time resolved spectra of C 480 in AOT reversed micelles, it was necessary to reinvestigate the photophysics of this dye in this media.C 480 is a water-soluble dye22 and its —uorescence efficiency (using Quinine sulfate in 0.05 M sulfuric acid as standard) and —uorescence lifetime in water are found to be 0.62 and 5.9 ns, respectively. Both these values are close to that reported earlier.22 Structured absorption and emission spectra are obtained in n-heptane and AOT reversed micelles when very low concentration of the dye is used (concentration D5]10~6 M). The absorption and emission spectra of the dye are given in Fig. 2 and 3, respectively. The following are some of the important observations from our study. The absorption spectrum of C 480 shows a broadening at the red edge on addition of 0.1 M AOT to nheptane. In addition to spectral broadening, an isosbestic point around 390 nm»where the absorbance of the dye in n-heptane is very low»is also observed with increase in W0 . The broadening at the red edge overlaps the absorption spectrum of the dye in water.A broadening of the emission spectrum in AOT binary solution is accompanied by a loss in structure as is evident from the ratio of the —uorescence intensity of the two peaks at 390 and 410 nm. A small red-shift, in overall, of the emission spectrum is also observed. Addition of water (increase in leads to an additional broadening of the W0) Fig. 2 Absorption spectrum of C 480 in n-heptane and AOT reversed micelles. The region where the isosbestic point is observed is circled. 5032 Phys. Chem. Chem. Phys., 1999, 1, 5029»5034Fig. 3 Steady-state —uorescence emission spectrum of C 480 in nheptane and AOT reversed micelles obtained on excitation at the respective longest wavelength of maximum absorption. The region where the isoemissive point is observed is circled. —uorescence spectrum and simultaneous appearance of a broad shoulder around 480 nm. In addition, a drastic quenching of —uorescence and an isoemissive point around 480 nm is also observed.It is important to note that the isoemissive point is observed in a region where the —uorescence of the dye in n-heptane is very small and that it is observed only on addition of water. The observations made in the absorption spectra suggest that in n-heptane»AOT binary solutions, the dye C 480 is partitioned between the oil and micellar phase. An isosbestic as well as isoemissive point observed on increasing shows W0 that the distribution of the dye molecules between the micellar phase and water molecules attains an equilibrium both in ground as well as excited states.28 Fluorescence lifetimes of C 480 were also determined for diÜerent excitation and emission wavelengths and for three diÜerent values of The –ndings are given in Table 1.It is W0 . evident from the table that the emission observed at 390 and 410 nm is contributed mainly by the dye molecules in nheptane. The decay at 490 nm is mono-exponential for W0\0 Table 1 Fluorescence lifetimes and pre-exponential factors obtained of C 480 in AOT reversed micelles for diÜerent values of and at W0 diÜerent excitation and detection wavelengths.a Excitation wavelength\360 nm, Detection wavelength\390 nm W0 ] 0 4 8 32 qF (ns) 2.34 2.22 2.28 2.27 A 1 1 1 1 Excitation wavelength\360 nm, Detection wavelength\410 nm W0 ] 0 4 8 32 qF (ns) 2.41 2.26 2.33 2.38 A 1 1 1 1 Excitation wavelength\360 nm, Detection wavelength\490 nm W0 qF1 (ns) A1 qF2 (ns) A2 0 4.74 1 4 1.8 [0.39 5.2 1.0 8 1.5 [0.28 5.35 1.0 32 0.97 [0.21 5.33 1.0 Excitation wavelength\420 nm, Detection wavelength\490 nm W0 ] 0 4 8 32 qF (ns) 5.26 5.17 5.26 A 1 1 1 a Lifetime of C 480 is 2.6 ns in n-heptane and 5.9 ns in water. only.For higher values a distinct rise time is observed. The W0 bi-exponential decays with a rise time are in agreement with the isoemissive point observed at 480 nm in the steady-state —uorescence spectra. A decrease in the rise time from 1.8 ns at to 0.97 ns at is observed.In contrast, the W0\4 W0\32 decay time in the bi-exponential decays is independent (around 5.3 ns) of suggesting that this corresponds to the W0 dye molecules solubilized in the free water pool of the reversed micelle. As in the case of BC I in AOT reversed micelles,11 we have been able to determine the —uorescence lifetime of the C 480 molecules absorbing in the red-edge of the absorption spectrum. Excitation of the dye at 420 nm and observation of the —uorescence at 490 nm gave mono-exponential decays at all values of The lifetime hence obtained was also around 5.3 W0 .ns and was almost independent of This value is close to W0 . the —uorescence lifetime of the dye in water22 (see also Table 1). This proves that the dye molecules absorbing in this region are the same as those observed on excitation at 360 nm Table 1), i.e., those that are present in the free water (qF2»see pool of the reversed micelles. Sarkar et al.7 suggest that the bi-exponential decays with a nanosecond rise time are due to the ììdecrease in the energy of the guest dipole with time due to solvation so that the solvated species emitting at longer wavelength is produced on the nanosecond scale.œœ The decay time of 7 ns reported by the same authors for the dye atW0\ should be taken with caution since the —uorescence lifetime 4 of C 480 is known to gradually increase with increase in solvent polarity and reach a value of 5.9 ns in water.22 Under the circumstances the dye molecules are present in two distinct regions (interface and water pool) of the reversed micelle.It is possible then, on selective excitation of the dye and on the assumption that there is no equilibrium between the concentration of the dye in the two diÜerent microenvironments, to probe using time correlated —uorescence spectroscopy the reorganization rates of the surrounding environment that is perturbed by electronic excitation of the probe dye.28 In view of the arguments presented above, it is likely that the time-resolved spectra of C 480 in AOT reversed micelles obtained in ref. 7 and by us in the present study are representative of dye molecules in the AOT micellar interface as well as in the water pool of the reversed micelles. Thus the 1.7 ns relaxation time with a pre-exponential factor of 0.5, observed at is perhaps not a true representation of the solva- W0\32 tion dynamics of the water pool which is a solvent property.This could be associated with the diÜusion of the probe dye between the micellar interface and water pool within its excited state lifetime which is a solute property, rather than exclusively to the solvation time of the free water molecules present in the water pool of the reversed micelle. Since we have shown that it is possible to determine the —uorescence lifetime of the C 480 molecules absorbing in the red-edge of the absorption spectrum, it will be worthwhile and interesting to study the solvation dynamics of C 480 in AOT reversed micelles by exciting the dye at 420 nm and using picosecond —uorescence spectrometers.The data analysis will be more trivial due to the non-participation of the chromophores present in n-heptane or interface in the resulting time resolved emission spectra. Conclusion In conclusion, nanosecond time resolved spectra of BC I and C 480 studied in AOT reversed micelles give diÜerent results for a water insoluble dye (BC I) and a water soluble dye (C 480).The relaxation rates obtained from the study using BC I may either be due to diÜusion of the probe dye from the apolar region to the micellar interface or to the solvent relaxation times of the bound water molecules. The solvent relaxation times obtained using C 480 as a probe more likely Phys. Chem. Chem. Phys., 1999, 1, 5029»5034 5033re—ect the diÜusion times of the probe dye between the interface and free water core rather than solvent relaxation of the free water molecules present in the pool.Acknowledgements work was supported by project 2/2.1/QUI/443/94. We This are extremely thankful to Dr. Alex Siemiarczuk (PTI, Canada) for his valuable suggestions in an efficient use of the timeresolved emission spectra module in the software provided with the instrument. BBR acknowledges PRAXIS XXI for the research grant (BPD/3993/96). We are also thankful to Dr. S. Pal and Prof. S. Seshadri for the kind gift of the dyes.References 1 M. Maroncelli, J. Mol. L iq., 1993, 57, 1. 2 S. Vajda, R. Jimenez, S. J. Rosenthal, V. Fidler, G. R. Fleming and E. W. Castner, Jr., J. Chem. Soc., Faraday T rans., 1995, 91, 867. 3 G. Saielli, A. Polimeno, P. L. Nordio, P. Bartolini, M. Ricci and R. Righini, J. Chem. Soc., Faraday T rans., 1998, 94, 121. 4 (a) N. Sarkar, D. Datta, S. Das and K. Bhattacharyya, J. Phys. Chem., 1996, 100, 15483. (b) S. Matzinger, D. M. Hussey and M.D. Feyer, J. Phys. Chem. B., 1998, 102, 7216. 5 A. Datta, S. K. Pal, D. Mandal and K. Bhattacharyya, J. Phys. Chem. B., 1998, 102, 6114. 6 K. Das, N. Sarkar, S. Das, A. Datta and K. Bhattacharyya, Chem. Phys. L ett., 1996, 249, 323. 7 N. Sarkar, K. Das, A. Datta, S. Das and K. Bhattacharyya, J. Phys. Chem., 1996, 100, 10523. 8 R. E. Riter, D. M. Willard and N. E. Levinger, J. Phys. Chem., 1998, 102B, 2705. 9 J. Zhang and F. V. Bright, J. Phys. Chem., 1991, 95, 7900. 10 N.Wittouck, R. M. Negri, M. Ameloot and F. C. De Shryver, J. Amer. Chem. Soc., 1994, 116, 10601. 11 B. B. Raju and S. M. B. Costa, J. Phys. Chem. B., 1999, 103B, 4309. 12 S. Pal, PhD. Thesis, Bombay University, 1991. 13 D. V. OœConnor and D. Phillips, in T ime-correlated single photon counting. Academic Press, New York, 1984, ch. 6. 14 M. Maroncelli and G. R. Fleming, J. Chem. Phys., 1987, 86, 6221. 15 B. B. Raju, J. Phys. Chem., 1997, 101A, 981. 16 B. B. Raju and S. M. B. Costa, Phys.Chem. Chem. Phys., 1999, 1, 3539. 17 B. B. Raju and B. Eliason, J. Photochem. Photobiol., A: Chem., 1988, 116, 135. 18 H. Shirota and K. Horie, J. Phys. Chem., 1999, 103B, 1437. 19 P. E. Zinsli, J. Phys. Chem., 1975, 83, 3223. 20 The authors in ref. 8 report that the mole fraction of intrinsic water to AOT as determined by Karl-Fischer titration, for (W0), the AOT sample used to be 1. The results reported for in W0\0 ref. 8 then should actually correspond to the results at W0\1 and that at to the results obtained at W0\2.8 W0\3.8. 21 W. Jarzeba, G. C. Walker, A. E. Johnson an P. F. Barbara, Chem. Phys., 1991, 152, 57. 22 G. Jones, II ; C. Y. Choi, W. R. Jackson and W. R. Bergmark, J. Phys. Chem., 1974, 89, 294. 23 H. Cho, M. Chung, J. Lee, T. Nguyn, S. Singh, M. Vedamuthu, S. Yao, S.-B. Zhu and G. W. Robinson, J. Phys. Chem., 1995, 99, 7806. 24 M. Bellete� te, M. Lachaoelle and G. Durocher, J. Phys. Chem., 1990, 94, 5337. 25 H. L. Pasquier, R. B.Pansu, J.-P. Chauvet, P. Pernot, A. Collet and J. Faure, L angmuir, 1997, 13, 1907. 26 W. Rettig and M. Zander, Ber. Bunsen-Ges. Phys. Chem., 1987, 87, 1143. 27 N. Mataga, H. Yao, T. Okada and W. Rettig, J. Phys. Chem., 1989, 93, 3383. 28 S. M. B. Costa, M. M. Velaç zquez, N. Tamai and I. Yamazaki, J. L umin., 1991, 48 & 49, 341 and references therein. Paper 9/06191F 5034 Phys. Chem. Chem. Phys., 1999, 1, 5029»5034 Nanosecond time resolved emission spectroscopy of aminocoumarins in AOT reversed micelles B.Bangar Raju§ and Silvia M. B. Costa* Centro de Quimica Estrutural, Complexo I, Instituto Superior T ecnico, Av. Rovisco Paris 1, 1049-001, L isboa Codex, Portugal. E-mail : FAX:]351-1-846 44 55 sbcosta=alfa.ist.utl.pt, Received 30th July 1999, Accepted 10th September 1999 Time resolved emission spectroscopy of a water insoluble aminocoumarin derivative, BC I, in n-heptane»AOT»water reversed micelles is investigated using a nanosecond —uorescence spectrometer.The extent of the time-dependent Stokes shift is found to be smaller for than for and also for W0\4 W0\10 molar ratio of water and AOT). A monoexponential decay of the dynamic Stokes shift W0\40 (W0 , correlation function with a time constant of 4.5 ns is observed at At the time-correlation W0\4. W0\10 function shows a bi-exponential decay with a fast component of D600 ps and another of 3.3 ns comparable to the —uorescence lifetime of the dye. The time correlation function at shows an extremely slow W0\40 component of the solvation time which is at least an order of magnitude larger than the —uorescence lifetime of the probe dye.In addition a faster component with a solvation time of 2.15 ns is also observed. The solvation time observed is likely due to the diÜusion of the probe dye within the micellar interface or due to the reorganization of the water molecules bound to the polar head groups of AOT. Steady-state absorption and —uorescence, and time resolved —uorescence study of another aminocoumarin, C 480, give evidence of equilibrium between the distribution of the dye molecules between the interface and water pool. The time resolved shift of the emission spectra is more likely due to the diÜusion of the probe between the interface and water pool rather than reorientation of the water molecules (solvent relaxation) in the core of the reversed micelles as it has been reported earlier.Introduction Water is a major and an important constituent of biological membranes and determines the dynamics of biological processes.Solvation dynamics in restricted media is now being widely investigated in an eÜort to understand the ultrafast processes that take place in biological membranes. This was stimulated by the depth of information obtained about the relaxation times of pure solvents and solvent mixtures using picosecond (ps) and femtosecond (fs) laser spectroscopy tools.1 While the solvent relaxation time of bulk water is in the subpicoseconds, in microheterogeneous media it is shown to vary, depending on the nature of water (free or trapped/bound), between few picoseconds and nanoseconds (ns).2h8 Aminocoumarin and anilinonaphthalene sulfonate (ANS) dyes are some of the most widely used probes to study solvation dynamics in restricted heterogeneous media, such as cyclodextrins,2 liquid crystals,3 micelles,4 vesicles,5 zeolites6 and reversed micelles.7h9 Location of the probe and the surroundings it experiences constitute an important aspect in —uorescence probe studies, whether steady-state or timeresolved, of microheterogeneous systems.For instance in Aerosol OT, AOT [sodium 1,4-bis(2-ethylhexyl) sulfosuccinate], reversed micelles the solubility of the dye and its preferential solvation, if any, at a speci–c site determine its absorption and emission characteristics. When a dye is soluble in all the three pseudophases of the reversed micelles, the distribution and diÜusion of the dye between the three diÜerent microenvironments and the partition of the dye as well as the § Present address : Department of Chemistry, University of Connecticut, 55 North Eagleville Road, Storrs, Connecticut 06269-3060, USA.diÜusion of the water molecules, need to be taken into consideration. This is in addition to the overall motion of the reversed micelles and the wobbling and tumbling motion of the probe dye in the reversed micelles.10 The aminocoumarin and ANS dyes that have so far been used to probe solvation dynamics in n-heptane»AOT»water reversed micelles are soluble in water as well as in alkane» AOT binary solutions.Hence the study of solvation dynamics is complicated due to the diÜusion of water molecules and, sometimes, also of the probe dye, between the oil, interface and the water pool. It is essential, hence, to characterize the location of the probe and whether or not there is a partitioning of the dye in the reversed micelles before extending the investigations to determine the time evolution of the —uorescence spectrum.We have attempted to investigate solvation dynamics of a water insoluble aminocoumarin dye, BC I, (Scheme 1) in nheptane »AOT»water reversed micelles. BC I is a non-rigid aminocoumarin probe dye with a —exible diethylamino and a cyano group as substituents at the 7- and 4-positions, respectively, of the parent coumarin.A second coumarin moiety is also present as a substituent at the 3-position. From the measurements of —uorescence excitation spectra at diÜerent emission wavelengths, steady-state —uorescence emission anisotropy and —uorescence decays at diÜerent excitation and emission wavelengths, it was recently shown by us11 that the dye experiences two diÜerent environments within the micellar interface. The dye is strongly in—uenced»due to hydrogen bonding»by the water molecules bound to the polar head groups of the surfactant.It was also shown that it is possible to selectively excite only those dye molecules that are located in the proximity of the polar head groups.11 Phys. Chem. Chem. Phys., 1999, 1, 5029»5034 5029 This journal is The Owner Societies 1999 (Scheme 1 Structures of the two 7-diethylaminocoumarin dyes discussed in the presented study. In this paper we discuss the time resolved —uorescence (nanosecond resolution) of BC I in the interfacial region of AOT reversed micelles for diÜerent values of W0 (W0\ We also revisit the reported solvation [water]/[AOT].dynamics of a water-soluble rigid aminocoumarin dye, C 480,7 (Scheme 1). A critical examination of the —uorescence spectrum of C 480 in AOT reversed micelles given as Fig. 2 in ref. 7 shows an isoemissive point in the region where there is no contribution of the —uorescence of the dye molecules in nheptane. This then implies that there is an equilibrium between the distribution of the dye molecules in the interfacial region and water pool of the reversed micelles.We hope to show that the results thus obtained can be better expined by taking into account the partition of this probe between the water and the interface rather than solvent relaxation, as reported earlier.7 Experimental The synthesis of BC I (a cyano derivative of coumarinyl benzopyrano pyridine) has been reported earlier.12 This was later puri–ed using column chromatography (silica gel, dichloromethane»ethylacetate). The purity of the dye has been veri–ed by thin layer chromatography on silica gel with the above solvent mixture as carrier.AOT is ultra pure grade from Sigma and n-heptane of spectroscopy grade is from Aldrich. Coumarin 480 is from Exciton. The three chemicals have been used as received. Water used was doubly distilled. The mole fraction of intrinsic water to AOT as deter- (W0), mined by Karl-Fischer titration, was found to be 0.2.Absorption spectra were recorded using Jasco V 500 UV/VIS/NIR absorption spectrometer. Steady-state —uorescence spectra were measured on Perkin Elmer 3B spectrometer with the cuvette holder placed in a perpendicular geometry. The spectral correction –le supplied by the manufacturer was used to obtain the corrected —uorescence spectrum. All the measurements were done at room temperature. Fluorescence lifetimes were determined with a PTI LS-1 (qf) time correlated single photon counting instrument using a nanosecond hydrogen —ash lamp.13 Fluctuations in the pulse jitter and intensity were corrected by making an alternate collection of scattering and sample emissions.In this way 10% of the total number of counts at the maximum were collected each time. Data analysis was carried out using the curve- –tting program supplied by the manufacturer. Reduced chisquared, s2, (0.95\s2\1.1) and a high Durbin»Watson parameter ([1.7) were considered while determining the quality of the –t.13 The error in the determination of the —uorescence lifetimes is ^200 ps.A utility in the software used to control the instrument is applied to generate —uorescence decays at diÜerent wavelengths and also to determine the maximum intensity of the decay at each measured wavelength. Spectrum reconstruction is performed following the method proposed by Maroncelli and Fleming.14 Brie—y, —uorescence decays are collected between 560 and 650 nm and in intervals of 10 nm.The data are then –t to a sum of exponentials. Then using the pre-exponential factors and the lifetimes, deconvoluted decays at the measured wavelengths are reconstructed. As the time-correlated single photon counting data is collected at a constant signal-to-noise ratio, the calculation of the integrated emission intensity is accomplished by scaling to the data acquisition time at each wavelength since this time is directly proportional to the integrated emission intensity.î Peak frequencies are determined by –tting the spectral points at the times of interest to a log-normal line shape function.14 The —uorescence maximum found in the steady-state spectrum is taken as the peak frequency at in–nite time.Results and discussion (a) Emission spectra of BC I at the AOT interface The spectroscopic characteristics of BC I have recently been investigated mainly in neat aprotic solvents15 and in dioxane» water mixtures.16 These have been explained using the intramolecular charge transfer (ICT) and the twisted ICT (TICT) hypothesis.Studies in dioxane»water mixtures also give evidence of a weak intermolecular hydrogen bond between the cyano nitrogen and water in the ground state. The dye is also found to show considerable —uorescence quenching with increase in the concentration of water in the solvent mixture. The non-radiative deactivation rates have been correlated with the rate of intramolecular charge transfer processes in the dye.Austin Model 1 (AM 1) semiempirical calculations recently done on another dye belonging to the same family show that in such dyes the ground state is nearly isoenergetic (S0) whether the second coumarin moiety is in-plane or twisted out-of-plane with the parent coumarin.17 The energy of the lowest excited state is however found to be lowest when (S1) the bulky substitution containing a second coumarin is coplanar to the parent coumarin containing the diethylamine and the cyano groups thus extending the p-electron conjugation in the excited singlet state.17 Such a probability, in addition to the dependence of the photophysics of BC I on the concentration of water in dioxane,16 thus, make an interesting study of the intramolecular charge transfer processes of the dye in con- –ned systems, such as reversed micelles.11 The location of the probe dye, BC I, and its photophysical properties in reversed micelles were investigated recently using steady-state and time dependent —uorescence techniques.The –ndings are summarized as follows :11 BC I, a hydrophobic aminocoumarin derivative, is preferentially solvated in the organic phase of the reversed micelle. However, in view of its size it experiences a multitude of microheterogeneous environments. In the reversed micelles, emission arising from two diÜerent environments for the dye is observed. In n-heptane BC I shows a partly structured absorption spectrum in the visible region.Addition of 0.1 M of AOT to n-heptane solution of BC I results, among other changes, in a signi–cant broadening of the absorption spectrum at the rededge (j[520 nm) region where the dye shows practically no absorption in n-heptane solutions. On addition of water to the reversed micelles further broadening, albeit small, of the absorption spectra at the red-edge and a decrease in the absorbance around 470 nm was observed. Selective excitation î As described in LS-100TM —uorescence lifetime analysis PTI} module reference manual 5030 Phys.Chem. Chem. Phys., 1999, 1, 5029»5034of the dye molecules absorbing at a wavelength greater than 500 nm resulted in a single broad —uorescence emission with a peak at about 572 nm for The —uorescence efficiency W0\0. was found to decrease with increase in A bathochromic W0 . shift of about 25 nm was also observed on increasing No W0 . —uorescence was observed on excitation of BC I around this wavelength ([500 nm) in normal n-heptane solutions.This implies that the —uorescence spectrum hence obtained is contributed by BC I molecules that are solubilized in the interfacial region of the reversed micelle and that absorb at wavelengths greater than 520 nm. Fluorescence decays obtained for excitation at 540 nm were all mono-exponential, varying from 4.2 ns at to 3.0 nm at The W0\4 W0\40. steady-state —uorescence emission anisotropy of BC I, carried out at an excitation wavelength of 540 nm gives a non-zero value11 (mean value) which increases non-linearly with increase in These results give evidence that the dye mol- W0 .ecules solubilized in the interface of the reversed micelle contribute to the emission observed on excitation at 540 nm. The quenching of the —uorescence quantum efficiency and lifetime of the dye with increase in could be attributed to the W0 hydrogen bonding between the cyano group and the water molecules (bound water) in the micellar region of the interface. 11 (b) Time resolved emission spectra Time resolved —uorescence spectra of BC I in reversed micelles (at 10 and 40) are constructed by exciting the dye W0\4, at 540 nm. At time zero, peak wavelength observed is D580 nm, close to the maximum value in the steady-state spectrum for For the overall time resolved shift is only W0\2. W0\4 about 156 wavenumbers while it was found to be almost double (327 wavenumbers) for The analysis for W0\10.is complicated due to the presence of an extremely W0\40 slow component in the time correlation function. In order to make a quantitative analysis of the time resolved spectra in the reversed micelles the Stokes shift correlation function C(t) was calculated using the expression :14 C(t)\ l(t)[l(O) l(0)[l(O) (1) where, l(t), l(O) and l(0) are, respectively, the emission maxima at time t, in–nity (fully relaxed) and zero. The timeresponse functions of the dynamic Stokes shift observed at and 10 are shown in Fig. 1a. The best –ts obtained are W0\4 given as solid lines. At the mono-exponential –t of the W0\4, correlation function gives a time constant of 4.54 ns. At W0\ there is an initial fast component (around 600 ps) and a 10 much slower one of 3.3 ns. The latter value is close to the excited state lifetime of BC I at The corresponding W0\10. time evolution of the —uorescence spectra is shown in Fig. 1b. The time correlation function for does not relax to W0\40 zero value within the time frame of the experimental observation (Fig. 1c). It could however be –t with a biexponential function with solvation times (pre-exponential factors) of 2.15 ns (0.80) and 21.3 ns (0.2). We could not resolve any extremely short-lived component, even if it were to be present, in the solvent relaxation function at this value of A long-lived W0 . second component of the solvation time has also been observed, earlier by Sarkar et al.7 and by us in this study (see below) for a rigid aminocoumarin dye, C 480, in aqueous reversed micelles.More recently Shirota and Horie18 have also observed a similar slow solvent relaxation time in nonaqueous reversed micelles using a rigid aminocoumarin dye, C 343. This long-lived component is generally found to be much larger than the —uorescence lifetime of the dye. With increase in although the viscosity of the reversed micelle decreases, W0 , the dielectric constant and, more importantly, the rotational relaxation time of the micelle increases18 and becomes (qrM) Fig. 1 (a) Time-response function of the dynamic Stokes shift observed at and 10. (b) Time resolved emission spectra of BC W0\4 I in AOT reversed micelles at (c) Time correlation function W0\10. for W0\40. much longer than the rotational diÜusion of the probe dye in the reversed micelle. This could be one reason for the large oÜset observed in the C(t) function. This shows up as a dark or non-observable solvation dynamics.Shirota and Horie18 observed that the dark solvation dynamics component increases with increase in This is in agreement with the W0 . Phys. Chem. Chem. Phys., 1999, 1, 5029»5034 5031results obtained earlier by Zinsli19 using Smoluchowsky diÜusion equations. The reasons for the large oÜset leading to a longer second component in the time correlation function in certain microheterogeneous systems are, however, not well understood. Zhang and Bright9 have used a charge transfer probe (ANS) to investigate the nanosecond reorganization rates of water molecules using frequency-domain —uorescence spectroscopy.They found that the reorganization time of bound water molecules reduces from D16 ns at to 3.5 ns at W0\0 W0\2.8.20 The relaxation times in the reversed micelles at and 10 W0\4 obtained by us in the present study using BC I as the probe dye are close to those reported by Zhang and Bright for For free water in the water pool, the same authors W0\2.8.9 report9 that the relaxation time decreases from 1.08 ns at to D500 ps at These values are much faster W0\0 W0\8.3.than the 8 ns and 1.7 ns times obtained, respectively, at W0\ and by Sarkar et al.7 using a structurally rigid 4 W0\32 Coumarin C 480. The average relaxation time SqT“ obtained by Sarkar et al.7 at (6.75 ns) is almost the same as W0\32 that at (8 ns). This implies that the solvent relaxation W0\4 does not become faster with increase in This is contrary W0 .to what one observes in reversed micelles.8,9 Solvent relaxation times reported by Sarkar et al. could then be due to some processes other than solvation dynamics. This subject is discussed in section c. The faster decay of the time-correlation function at W0\40 and 10 than at obtained in the present study is consis- W0\4 tent with the reorganization times obtained by Zang and Bright,9 Vajda et al.2 using ICT probes. Recently, the latter authors2 have found that the solvation dynamics of Coumarin 460 (a TICT candidate-dye) in c-cyclodextrin is faster than that of the rigid aminocoumarin dye, C 480.Barbara and coworkers also observed that the dielectric relaxation times of the solvents was faster when observed using Coumarin 120 (7- dimethylamino-4-tri—uoromethyl coumarin) as the probe dye.21 In neat solvents, the solvation dynamics of C 120»a dye that is reported to undergo TICT in highly polar solvents21,22»has been reported to be 20 to 40% faster than the structurally rigid aminocoumarin dye, C 153.21 In reversed micelles ICT processes (correlated to the decrease in the —uorescence lifetime and increase in rate of non-radiative decay) are known to become faster with increase in W0 values.23 The polarity of the interior of the reversed micelles increases with increase in from 4 to 10.23,24 Increase in W0 polarity of the solvent medium is known to facilitate a greater charge transfer in BC I.15,16 The decrease in the —uorescence lifetime with increase in is due to the increase in W0 ICT.11,15,16 This speeds up the reorganization of the solvent around the newly formed solute dipoles in the excited state.The solvation time of BC I in reversed micelles at W0\40 and 10 is, however, three orders of magnitude slower than that observed in bulk water.8 As mentioned earlier, no isosbestic or isoemissive point is observed in the absorption and —uorescence spectra of BC I for diÜerent values of in AOT reversed micelles.In view of W0 its hydrophobic nature the dye resides solely in the oil and in the micellar interface. To our knowledge there has been only one earlier report of time-resolved —uorescence study of a dye present exclusively in the micellar interface. Pansu and coworkers25 studied the diÜusion kinetics of bianthryls (BA and BOA) in CTAC micelles in the excited state. Bianthryl, BA, is a TICT candidate dye26,27 and is a neutral molecule and insoluble in water.It is hence solubilized in the core of the hydrophobic domain of the micelle. The time-dependent Stokes shift of the —uorescence spectra showed two temporary isoemissive points for BA and one for BOA. The dynamic where A is the pre-exponential factor and q is “ SqT\A1q1]A2q2 , the lifetime. Stokes shift has been interpreted as due to a ììdiÜusion of the excited probe from the non-polar hydrophobic core to the water interface of the micelle where it is eventually trappedœœ. BC I is also a neutral molecule and is insoluble in water.In reversed micelles it shows a preferential solvation in the oil phase. Thus the relaxation times observed in the present study may either be due to the diÜusion of the dye molecules within the interface or to the reorganization of the highly constrained bound water molecules in the interface. (c) Study of C 480 in AOT reversed micelles We have also investigated the time resolved spectra of C 480 in AOT reversed micelles and obtained the time resolved emission spectrum after subtracting the dye molecules contribution in n-heptane.The relaxation times (1.7 ns and 12 ns) and the pre-exponential factors (0.5 and 0.5, respectively) obtained by us for C(t) using a nanosecond —uorescence spectrophotometer agree with those reported by Sarkar et al.7 for C 480 at Although the 1.7 ns component is within W0\32. the —uorescence lifetime of the dye molecule, the physical meaning of the dominant 12 ns component is not trivial to comprehend.It is more likely to be representative of a process associated with the translational or rotational reorientation or diÜusion time of the reversed micelle rather than solvation dynamics of the probe dye.19 In order to understand the time resolved spectra of C 480 in AOT reversed micelles, it was necessary to reinvestigate the photophysics of this dye in this media. C 480 is a water-soluble dye22 and its —uorescence efficiency (using Quinine sulfate in 0.05 M sulfuric acid as standard) and —uorescence lifetime in water are found to be 0.62 and 5.9 ns, respectively.Both these values are close to that reported earlier.22 Structured absorption and emission spectra are obtained in n-heptane and AOT reversed micelles when very low concentration of the dye is used (concentration D5]10~6 M). The absorption and emission spectra of the dye are given in Fig. 2 and 3, respectively.The following are some of the important observations from our study. The absorption spectrum of C 480 shows a broadening at the red edge on addition of 0.1 M AOT to nheptane. In addition to spectral broadening, an isosbestic point around 390 nm»where the absorbance of the dye in n-heptane is very low»is also observed with increase in W0 . The broadening at the red edge overlaps the absorption spectrum of the dye in water. A broadening of the emission spectrum in AOT binary solution is accompanied by a loss in structure as is evident from the ratio of the —uorescence intensity of the two peaks at 390 and 410 nm.A small red-shift, in overall, of the emission spectrum is also observed. Addition of water (increase in leads to an additional broadening of the W0) Fig. 2 Absorption spectrum of C 480 in n-heptane and AOT reversed micelles. The region where the isosbestic point is observed is circled. 5032 Phys. Chem. Chem. Phys., 1999, 1, 5029»5034Fig. 3 Steady-state —uorescence emission spectrum of C 480 in nheptane and AOT reversed micelles obtained on excitation at the respective longest wavelength of maximum absorption. The region where the isoemissive point is observed is circled. —uorescence spectrum and simultaneous appearance of a broad shoulder around 480 nm. In addition, a drastic quenching of —uorescence and an isoemissive point around 480 nm is also observed. It is important to note that the isoemissive point is observed in a region where the —uorescence of the dye in n-heptane is very small and that it is observed only on addition of water.The observations made in the absorption spectra suggest that in n-heptane»AOT binary solutions, the dye C 480 is partitioned between the oil and micellar phase. An isosbestic as well as isoemissive point observed on increasing shows W0 that the distribution of the dye molecules between the micellar phase and water molecules attains an equilibrium both in ground as well as excited states.28 Fluorescence lifetimes of C 480 were also determined for diÜerent excitation and emission wavelengths and for three diÜerent values of The –ndings are given in Table 1.It is W0 . evident from the table that the emission observed at 390 and 410 nm is contributed mainly by the dye molecules in nheptane. The decay at 490 nm is mono-exponential for W0\0 Table 1 Fluorescence lifetimes and pre-exponential factors obtained of C 480 in AOT reversed micelles for diÜerent values of and at W0 diÜerent excitation and detection wavelengths.a Excitation wavelength\360 nm, Detection wavelength\390 nm W0 ] 0 4 8 32 qF (ns) 2.34 2.22 2.28 2.27 A 1 1 1 1 Excitation wavelength\360 nm, Detection wavelength\410 nm W0 ] 0 4 8 32 qF (ns) 2.41 2.26 2.33 2.38 A 1 1 1 1 Excitation wavelength\360 nm, Detection wavelength\490 nm W0 qF1 (ns) A1 qF2 (ns) A2 0 4.74 1 4 1.8 [0.39 5.2 1.0 8 1.5 [0.28 5.35 1.0 32 0.97 [0.21 5.33 1.0 Excitation wavelength\420 nm, Detection wavelength\490 nm W0 ] 0 4 8 32 qF (ns) 5.26 5.17 5.26 A 1 1 1 a Lifetime of C 480 is 2.6 ns in n-heptane and 5.9 ns in water.only. For higher values a distinct rise time is observed. The W0 bi-exponential decays with a rise time are in agreement with the isoemissive point observed at 480 nm in the steady-state —uorescence spectra. A decrease in the rise time from 1.8 ns at to 0.97 ns at is observed. In contrast, the W0\4 W0\32 decay time in the bi-exponential decays is independent (around 5.3 ns) of suggesting that this corresponds to the W0 dye molecules solubilized in the free water pool of the reversed micelle.As in the case of BC I in AOT reversed micelles,11 we have been able to determine the —uorescence lifetime of the C 480 molecules absorbing in the red-edge of the absorption spectrum. Excitation of the dye at 420 nm and observation of the —uorescence at 490 nm gave mono-exponential decays at all values of The lifetime hence obtained was also around 5.3 W0 .ns and was almost independent of This value is close to W0 . the —uorescence lifetime of the dye in water22 (see also Table 1). This proves that the dye molecules absorbing in this region are the same as those observed on excitation at 360 nm Table 1), i.e., those that are present in the free water (qF2»see pool of the reversed micelles. Sarkar et al.7 suggest that the bi-exponential decays with a nanosecond rise time are due to the ììdecrease in the energy of the guest dipole with time due to solvation so that the solvated species emitting at longer wavelength is produced on the nanosecond scale. œœ The decay time of 7 ns reported by the same authors for the dye atW0\ should be taken with caution since the —uorescence lifetime 4 of C 480 is known to gradually increase with increase in solvent polarity and reach a value of 5.9 ns in water.22 Under the circumstances the dye molecules are present in two distinct regions (interface and water pool) of the reversed micelle.It is possible then, on selective excitation of the dye and on the assumption that there is no equilibrium between the concentration of the dye in the two diÜerent microenvironments, to probe using time correlated —uorescence spectroscopy the reorganization rates of the surrounding environment that is perturbed by electronic excitation of the probe dye.28 In view of the arguments presented above, it is likely that the time-resolved spectra of C 480 in AOT reversed micelles obtained in ref. 7 and by us in the present study are representative of dye molecules in the AOT micellar interface as well as in the water pool of the reversed micelles. Thus the 1.7 ns relaxation time with a pre-exponential factor of 0.5, observed at is perhaps not a true representation of the solva- W0\32 tion dynamics of the water pool which is a solvent property. This could be associated with the diÜusion of the probe dye between the micellar interface and water pool within its excited state lifetime which is a solute property, rather than exclusively to the solvation time of the free water molecules present in the water pool of the reversed micelle.Since we have shown that it is possible to determine the —uorescence lifetime of the C 480 molecules absorbing in the red-edge of the absorption spectrum, it will be worthwhile and interesting to study the solvation dynamics of C 480 in AOT reversed micelles by exciting the dye at 420 nm and using picosecond —uorescence spectrometers.The data analysis will be more trivial due to the non-participation of the chromophores present in n-heptane or interface in the resulting time resolved emission spectra. Conclusion In conclusion, nanosecond time resolved spectra of BC I and C 480 studied in AOT reversed micelles give diÜerent results for a water insoluble dye (BC I) and a water soluble dye (C 480).The relaxation rates obtained from the study using BC I may either be due to diÜusion of the probe dye from the apolar region to the micellar interface or to the solvent relaxation times of the bound water molecules. The solvent relaxation times obtained using C 480 as a probe more likely Phys. Chem. Chem. Phys., 1999, 1, 5029»5034 5033re—ect the diÜusion times of the probe dye between the interface and free water core rather than solvent relaxation of the free water molecules present in the pool.Acknowledgements work was supported by project 2/2.1/QUI/443/94. We This are extremely thankful to Dr. Alex Siemiarczuk (PTI, Canada) for his valuable suggestions in an efficient use of the timeresolved emission spectra module in the software provided with the instrument. BBR acknowledges PRAXIS XXI for the research grant (BPD/3993/96). We are also thankful to Dr. S. Pal and Prof. S. Seshadri for the kind gift of the dyes. References 1 M.Maroncelli, J. Mol. L iq., 1993, 57, 1. 2 S. Vajda, R. Jimenez, S. J. Rosenthal, V. Fidler, G. R. Fleming and E. W. Castner, Jr., J. Chem. Soc., Faraday T rans., 1995, 91, 867. 3 G. Saielli, A. Polimeno, P. L. Nordio, P. Bartolini, M. Ricci and R. Righini, J. Chem. Soc., Faraday T rans., 1998, 94, 121. 4 (a) N. Sarkar, D. Datta, S. Das and K. Bhattacharyya, J. Phys. Chem., 1996, 100, 15483. (b) S. Matzinger, D. M. Hussey and M. D. Feyer, J. Phys. Chem. B., 1998, 102, 7216. 5 A. Datta, S. K. Pal, D. Mandal and K. Bhattacharyya, J. Phys. Chem. B., 1998, 102, 6114. 6 K. Das, N. Sarkar, S. Das, A. Datta and K. Bhattacharyya, Chem. Phys. L ett., 1996, 249, 323. 7 N. Sarkar, K. Das, A. Datta, S. Das and K. Bhattacharyya, J. Phys. Chem., 1996, 100, 10523. 8 R. E. Riter, D. M. Willard and N. E. Levinger, J. Phys. Chem., 1998, 102B, 2705. 9 J. Zhang and F. V. Bright, J. Phys. Chem., 1991, 95, 7900. 10 N. Wittouck, R. M. Negri, M. Ameloot and F. C. De Shryver, J.Amer. Chem. Soc., 1994, 116, 10601. 11 B. B. Raju and S. M. B. Costa, J. Phys. Chem. B., 1999, 103B, 4309. 12 S. Pal, PhD. Thesis, Bombay University, 1991. 13 D. V. OœConnor and D. Phillips, in T ime-correlated single photon counting. Academic Press, New York, 1984, ch. 6. 14 M. Maroncelli and G. R. Fleming, J. Chem. Phys., 1987, 86, 6221. 15 B. B. Raju, J. Phys. Chem., 1997, 101A, 981. 16 B. B. Raju and S. M. B. Costa, Phys. Chem. Chem. Phys., 1999, 1, 3539. 17 B.B. Raju and B. Eliason, J. Photochem. Photobiol., A: Chem., 1988, 116, 135. 18 H. Shirota and K. Horie, J. Phys. Chem., 1999, 103B, 1437. 19 P. E. Zinsli, J. Phys. Chem., 1975, 83, 3223. 20 The authors in ref. 8 report that the mole fraction of intrinsic water to AOT as determined by Karl-Fischer titration, for (W0), the AOT sample used to be 1. The results reported for in W0\0 ref. 8 then should actually correspond to the results at W0\1 and that at to the results obtained at W0\2.8 W0\3.8. 21 W. Jarzeba, G. C. Walker, A. E. Johnson an P. F. Barbara, Chem. Phys., 1991, 152, 57. 22 G. Jones, II ; C. Y. Choi, W. R. Jackson and W. R. Bergmark, J. Phys. Chem., 1974, 89, 294. 23 H. Cho, M. Chung, J. Lee, T. Nguyn, S. Singh, M. Vedamuthu, S. Yao, S.-B. Zhu and G. W. Robinson, J. Phys. Chem., 1995, 99, 7806. 24 M. Bellete� te, M. Lachaoelle and G. Durocher, J. Phys. Chem., 1990, 94, 5337. 25 H. L. Pasquier, R. B. Pansu, J.-P. Chauvet, P. Pernot, A. Collet and J.Faure, L angmuir, 1997, 13, 1907. 26 W. Rettig and M. Zander, Ber. Bunsen-Ges. Phys. Chem., 1987, 87, 1143. 27 N. Mataga, H. Yao, T. Okada and W. Rettig, J. Phys. Chem., 1989, 93, 3383. 28 S. M. B. Costa, M. M. Velaç zquez, N. Tamai and I. Yamazaki, J. L umin., 1991, 48 & 49, 341 and references therein. Paper 9/06191F 5034 Phys. Chem. Chem. Phys., 1999, 1, 5029»5034 Nanosecond time resolved emission spectroscopy of aminocoumarins in AOT reversed micelles B. Bangar Raju§ and Silvia M.B. Costa* Centro de Quimica Estrutural, Complexo I, Instituto Superior T ecnico, Av. Rovisco Paris 1, 1049-001, L isboa Codex, Portugal. E-mail : FAX:]351-1-846 44 55 sbcosta=alfa.ist.utl.pt, Received 30th July 1999, Accepted 10th September 1999 Time resolved emission spectroscopy of a water insoluble aminocoumarin derivative, BC I, in n-heptane»AOT»water reversed micelles is investigated using a nanosecond —uorescence spectrometer. The extent of the time-dependent Stokes shift is found to be smaller for than for and also for W0\4 W0\10 molar ratio of water and AOT).A monoexponential decay of the dynamic Stokes shift W0\40 (W0 , correlation function with a time constant of 4.5 ns is observed at At the time-correlation W0\4. W0\10 function shows a bi-exponential decay with a fast component of D600 ps and another of 3.3 ns comparable to the —uorescence lifetime of the dye. The time correlation function at shows an extremely slow W0\40 component of the solvation time which is at least an order of magnitude larger than the —uorescence lifetime of the probe dye.In addition a faster component with a solvation time of 2.15 ns is also observed. The solvation time observed is likely due to the diÜusion of the probe dye within the micellar interface or due to the reorganization of the water molecules bound to the polar head groups of AOT. Steady-state absorption and —uorescence, and time resolved —uorescence study of another aminocoumarin, C 480, give evidence of equilibrium between the distribution of the dye molecules between the interface and water pool.The time resolved shift of the emission spectra is more likely due to the diÜusion of the probe between the interface and water pool rather than reorientation of the water molecules (solvent relaxation) in the core of the reversed micelles as it has been reported earlier. Introduction Water is a major and an important constituent of biological membranes and determines the dynamics of biological processes.Solvation dynamics in restricted media is now being widely investigated in an eÜort to understand the ultrafast processes that take place in biological membranes. This was stimulated by the depth of information obtained about the relaxation times of pure solvents and solvent mixtures using picosecond (ps) and femtosecond (fs) laser spectroscopy tools.1 While the solvent relaxation time of bulk water is in the subpicoseconds, in microheterogeneous media it is shown to vary, depending on the nature of water (free or trapped/bound), between few picoseconds and nanoseconds (ns).2h8 Aminocoumarin and anilinonaphthalene sulfonate (ANS) dyes are some of the most widely used probes to study solvation dynamics in restricted heterogeneous media, such as cyclodextrins,2 liquid crystals,3 micelles,4 vesicles,5 zeolites6 and reversed micelles.7h9 Location of the probe and the surroundings it experiences constitute an important aspect in —uorescence probe studies, whether steady-state or timeresolved, of microheterogeneous systems.For instance in Aerosol OT, AOT [sodium 1,4-bis(2-ethylhexyl) sulfosuccinate], reversed micelles the solubility of the dye and its preferential solvation, if any, at a speci–c site determine its absorption and emission characteristics. When a dye is soluble in all the three pseudophases of the reversed micelles, the distribution and diÜusion of the dye between the three diÜerent microenvironments and the partition of the dye as well as the § Present address : Department of Chemistry, University of Connecticut, 55 North Eagleville Road, Storrs, Connecticut 06269-3060, USA.diÜusion of the water molecules, need to be taken into consideration. This is in addition to the overall motion of the reversed micelles and the wobbling and tumbling motion of the probe dye in the reversed micelles.10 The aminocoumarin and ANS dyes that have so far been used to probe solvation dynamics in n-heptane»AOT»water reversed micelles are soluble in water as well as in alkane» AOT binary solutions.Hence the study of solvation dynamics is complicated due to the diÜusion of water molecules and, sometimes, also of the probe dye, between the oil, interface and the water pool. It is essential, hence, to characterize the location of the probe and whether or not there is a partitioning of the dye in the reversed micelles before extending the investigations to determine the time evolution of the —uorescence spectrum.We have attempted to investigate solvation dynamics of a water insoluble aminocoumarin dye, BC I, (Scheme 1) in nheptane »AOT»water reversed micelles. BC I is a non-rigid aminocoumarin probe dye with a —exible diethylamino and a cyano group as substituents at the 7- and 4-positions, respectively, of the parent coumarin. A second coumarin moiety is also present as a substituent at the 3-position. From the measurements of —uorescence excitation spectra at diÜerent emission wavelengths, steady-state —uorescence emission anisotpy and —uorescence decays at diÜerent excitation and emission wavelengths, it was recently shown by us11 that the dye experiences two diÜerent environments within the micellar interface. The dye is strongly in—uenced»due to hydrogen bonding»by the water molecules bound to the polar head groups of the surfactant.It was also shown that it is possible to selectively excite only those dye molecules that are located in the proximity of the polar head groups.11 Phys. Chem. Chem. Phys., 1999, 1, 5029»5034 5029 This journal is The Owner Societies 1999 (Scheme 1 Structures of the two 7-diethylaminocoumarin dyes discussed in the presented study. In this paper we discuss the time resolved —uorescence (nanosecond resolution) of BC I in the interfacial region of AOT reversed micelles for diÜerent values of W0 (W0\ We also revisit the reported solvation [water]/[AOT].dynamics of a water-soluble rigid aminocoumarin dye, C 480,7 (Scheme 1). A critical examination of the —uorescence spectrum of C 480 in AOT reversed micelles given as Fig. 2 in ref. 7 shows an isoemissive point in the region where there is no contribution of the —uorescence of the dye molecules in nheptane. This then implies that there is an equilibrium between the distribution of the dye molecules in the interfacial region and water pool of the reversed micelles.We hope to show that the results thus obtained can be better explained by taking into account the partition of this probe between the water and the interface rather than solvent relaxation, as reported earlier.7 Experimental The synthesis of BC I (a cyano derivative of coumarinyl benzopyrano pyridine) has been reported earlier.12 This was later puri–ed using column chromatography (silica gel, dichloromethane»ethylacetate).The purity of the dye has been veri–ed by thin layer chromatography on silica gel with the above solvent mixture as carrier. AOT is ultra pure grade from Sigma and n-heptane of spectroscopy grade is from Aldrich. Coumarin 480 is from Exciton. The three chemicals have been used as received. Water used was doubly distilled. The mole fraction of intrinsic water to AOT as deter- (W0), mined by Karl-Fischer titration, was found to be 0.2. Absorption spectra were recorded using Jasco V 500 UV/VIS/NIR absorption spectrometer. Steady-state —uorescence spectra were measured on Perkin Elmer 3B spectrometer with the cuvette holder placed in a perpendicular geometry.The spectral correction –le supplied by the manufacturer was used to obtain the corrected —uorescence spectrum. All the measurements were done at room temperature. Fluorescence lifetimes were determined with a PTI LS-1 (qf) time correlated single photon counting instrument using a nanosecond hydrogen —ash lamp.13 Fluctuations in the pulse jitter and intensity were corrected by making an alternate collection of scattering and sample emissions.In this way 10% of the total number of counts at the maximum were collected each time. Data analysis was carried out using the curve- –tting program supplied by the manufacturer. Reduced chisquared, s2, (0.95\s2\1.1) and a high Durbin»Watson parameter ([1.7) were considered while determining the quality of the –t.13 The error in the determination of the —uorescence lifetimes is ^200 ps.A utility in the software used to control the instrument is applied to generate —uorescence decays at diÜerent wavelengths and also to determine the maximum intensity of the decay at each measured wavelength. Spectrum reconstruction is performed following the method proposed by Maroncelli and Fleming.14 Brie—y, —uorescence decays are collected between 560 and 650 nm and in intervals of 10 nm. The data are then –t to a sum of exponentials.Then using the pre-exponential factors and the lifetimes, deconvoluted decays at the measured wavelengths are reconstructed. As the time-correlated single photon counting data is collected at a constant signal-to-noise ratio, the calculation of the integrated emission intensity is accomplished by scaling to the data acquisition time at each wavelength since this time is directly proportional to the integrated emission intensity.î Peak frequencies are determined by –tting the spectral points at the times of interest to a log-normal line shape function.14 The —uorescence maximum found in the steady-state spectrum is taken as the peak frequency at in–nite time.Results and discussion (a) Emission spectra of BC I at the AOT interface The spectroscopic characteristics of BC I have recently been investigated mainly in neat aprotic solvents15 and in dioxane» water mixtures.16 These have been explained using the intramolecular charge transfer (ICT) and the twisted ICT (TICT) hypothesis. Studies in dioxane»water mixtures also give evidence of a weak intermolecular hydrogen bond between the cyano nitrogen and water in the ground state. The dye is also found to show considerable —uorescence quenching with increase in the concentration of water in the solvent mixture.The non-radiative deactivation rates have been correlated with the rate of intramolecular charge transfer processes in the dye.Austin Model 1 (AM 1) semiempirical calculations recently done on another dye belonging to the same family show that in such dyes the ground state is nearly isoenergetic (S0) whether the second coumarin moiety is in-plane or twisted out-of-plane with the parent coumarin.17 The energy of the lowest excited state is however found to be lowest when (S1) the bulky substitution containing a second coumarin is coplanar to the parent coumarin containing the diethylamine and the cyano groups thus extending the p-electron conjugation in the excited singlet state.17 Such a probability, in addition to the dependence of the photophysics of BC I on the concentration of water in dioxane,16 thus, make an interesting study of the intramolecular charge transfer processes of the dye in con- –ned systems, such as reversed micelles.11 The location of the probe dye, BC I, and its photophysical properties in reversed micelles were investigated recently using steady-state and time dependent —uorescence techniques.The –ndings are summarized as follows :11 BC I, a hydrophobic aminocoumarin derivative, is preferentially solvated in the organic phase of the reversed micelle. However, in view of its size it experiences a multitude of microheterogeneous environments. In the reversed micelles, emission arising from two diÜerent environments for the dye is observed. In n-heptane BC I shows a partly structured absorption spectrum in the visible region.Addition of 0.1 M of AOT to n-heptane solution of BC I results, among other changes, in a signi–cant broadening of the absorption spectrum at the rededge (j[520 nm) region where the dye shows practically no absorption in n-heptane solutions. On addition of water to the reversed micelles further broadening, albeit small, of the absorption spectra at the red-edge and a decrease in the absorbance around 470 nm was observed. Selective excitation î As described in LS-100TM —uorescence lifetime analysis PTI} module reference manual 5030 Phys.Chem. Chem. Phys., 1999, 1, 5029»5034of the dye molecules absorbing at a wavelength greater than 500 nm resulted in a single broad —uorescence emission with a peak at about 572 nm for The —uorescence efficiency W0\0. was found to decrease with increase in A bathochromic W0 . shift of about 25 nm was also observed on increasing No W0 . —uorescence was observed on excitation of BC I around this wavelength ([500 nm) in normal n-heptane solutions.This implies that the —uorescence spectrum hence obtained is contributed by BC I molecules that are solubilized in the interfacial region of the reversed micelle and that absorb at wavelengths greater than 520 nm. Fluorescence decays obtained for excitation at 540 nm were all mono-exponential, varying from 4.2 ns at to 3.0 nm at The W0\4 W0\40. steady-state —uorescence emission anisotropy of BC I, carried out at an excitation wavelength of 540 nm gives a non-zero value11 (mean value) which increases non-linearly with increase in These results give evidence that the dye mol- W0 .ecules solubilized in the interface of the reversed micelle contribute to the emission observed on excitation at 540 nm. The quenching of the —uorescence quantum efficiency and lifetime of the dye with increase in could be attributed to the W0 hydrogen bonding between the cyano group and the water molecules (bound water) in the micellar region of the interface. 11 (b) Time resolved emission spectra Time resolved —uorescence spectra of BC I in reversed micelles (at 10 and 40) are constructed by exciting the dye W0\4, at 540 nm. At time zero, peak wavelength observed is D580 nm, close to the maximum value in the steady-state spectrum for For the overall time resolved shift is only W0\2. W0\4 about 156 wavenumbers while it was found to be almost double (327 wavenumbers) for The analysis for W0\10. is complicated due to the presence of an extremely W0\40 slow component in the time correlation function.In order to make a quantitative analysis of the time resolved spectra in the reversed micelles the Stokes shift correlation function C(t) was calculated using the expression :14 C(t)\ l(t)[l(O) l(0)[l(O) (1) where, l(t), l(O) and l(0) are, respectively, the emission maxima at time t, in–nity (fully relaxed) and zero. The timeresponse functions of the dynamic Stokes shift observed at and 10 are shown in Fig. 1a. The best –ts obtained are W0\4 given as solid lines. At the mono-exponential –t of the W0\4, correlation function gives a time constant of 4.54 ns. At W0\ there is an initial fast component (around 600 ps) and a 10 much slower one of 3.3 ns. The latter value is close to the excited state lifetime of BC I at The corresponding W0\10. time evolution of the —uorescence spectra is shown in Fig. 1b. The time correlation function for does not relax to W0\40 zero value within the time frame of the experimental observation (Fig. 1c). It could however be –t with a biexponential function with solvation times (pre-exponential factors) of 2.15 ns (0.80) and 21.3 ns (0.2). We could not resolve any extremely short-lived component, even if it were to be present, in the solvent relaxation function at this value of A long-lived W0 . second component of the solvation time has also been observed, earlier by Sarkar et al.7 and by us in this study (see below) for a rigid aminocoumarin dye, C 480, in aqueous reversed micelles.More recently Shirota and Horie18 have also observed a similar slow solvent relaxation time in nonaqueous reversed micelles using a rigid aminocoumarin dye, C 343. This long-lived component is generally found to be much larger than the —uorescence lifetime of the dye. With increase in although the viscosity of the reversed micelle decreases, W0 , the dielectric constant and, more importantly, the rotational relaxation time of the micelle increases18 and becomes (qrM) Fig. 1 (a) Time-response function of the dynamic Stokes shift observed at and 10. (b) Time resolved emission spectra of BC W0\4 I in AOT reversed micelles at (c) Time correlation function W0\10. for W0\40. much longer than the rotational diÜusion of the probe dye in the reversed micelle. This could be one reason for the large oÜset observed in the C(t) function. This shows up as a dark or non-observable solvation dynamics.Shirota and Horie18 observed that the dark solvation dynamics component increases with increase in This is in agreement with the W0 . Phys. Chem. Chem. Phys., 1999, 1, 5029»5034 5031results obtained earlier by Zinsli19 using Smoluchowsky diÜusion equations. The reasons for the large oÜset leading to a longer second component in the time correlation function in certain microheterogeneous systems are, however, not well understood. Zhang and Bright9 have used a charge transfer probe (ANS) to investigate the nanosecond reorganization rates of water molecules using frequency-domain —uorescence spectroscopy.They found that the reorganization time of bound water molecules reduces from D16 ns at to 3.5 ns at W0\0 W0\2.8.20 The relaxation times in the reversed micelles at and 10 W0\4 obtained by us in the present study using BC I as the probe dye are close to those reported by Zhang and Bright for For free water in the water pool, the same authors W0\2.8.9 report9 that the relaxation time decreases from 1.08 ns at to D500 ps at These values are much faster W0\0 W0\8.3.than the 8 ns and 1.7 ns times obtained, respectively, at W0\ and by Sarkar et al.7 using a structurally rigid 4 W0\32 Coumarin C 480. The average relaxation time SqT“ obtained by Sarkar et al.7 at (6.75 ns) is almost the same as W0\32 that at (8 ns). This implies that the solvent relaxation W0\4 does not become faster with increase in This is contrary W0 .to what one observes in reversed micelles.8,9 Solvent relaxation times reported by Sarkar et al. could then be due to some processes other than solvation dynamics. This subject is discussed in section c. The faster decay of the time-correlation function at W0\40 and 10 than at obtained in the present study is consis- W0\4 tent with the reorganization times obtained by Zang and Bright,9 Vajda et al.2 using ICT probes. Recently, the latter authors2 have found that the solvation dynamics of Coumarin 460 (a TICT candidate-dye) in c-cyclodextrin is faster than that of the rigid aminocoumarin dye, C 480.Barbara and coworkers also observed that the dielectric relaxation times of the solvents was faster when observed using Coumarin 120 (7- dimethylamino-4-tri—uoromethyl coumarin) as the probe dye.21 In neat solvents, the solvation dynamics of C 120»a dye that is reported to undergo TICT in highly polar solvents21,22»has been reported to be 20 to 40% faster than the structurally rigid aminocoumarin dye, C 153.21 In reversed micelles ICT processes (correlated to the decrease in the —uorescence lifetime and increase in rate of non-radiative decay) are known to become faster with increase in W0 values.23 The polarity of the interior of the reversed micelles increases with increase in from 4 to 10.23,24 Increase in W0 polarity of the solvent medium is known to facilitate a greater charge transfer in BC I.15,16 The decrease in the —uorescence lifetime with increase in is due to the increase in W0 ICT.11,15,16 This speeds up the reorganization of the solvent around the newly formed solute dipoles in the excited state.The solvation time of BC I in reversed micelles at W0\40 and 10 is, however, three orders of magnitude slower than that observed in bulk water.8 As mentioned earlier, no isosbestic or isoemissive point is observed in the absorption and —uorescence spectra of BC I for diÜerent values of in AOT reversed micelles.In view of W0 its hydrophobic nature the dye resides solely in the oil and in the micellar interface. To our knowledge there has been only one earlier report of time-resolved —uorescence study of a dye present exclusively in the micellar interface. Pansu and coworkers25 studied the diÜusion kinetics of bianthryls (BA and BOA) in CTAC micelles in the excited state. Bianthryl, BA, is a TICT candidate dye26,27 and is a neutral molecule and insoluble in water.It is hence solubilized in the core of the hydrophobic domain of the micelle. The time-dependent Stokes shift of the —uorescence spectra showed two temporary isoemissive points for BA and one for BOA. The dynamic where A is the pre-exponential factor and q is “ SqT\A1q1]A2q2 , the lifetime. Stokes shift has been interpreted as due to a ììdiÜusion of the excited probe from the non-polar hydrophobic core to the water interface of the micelle where it is eventually trappedœœ.BC I is also a neutral molecule and is insoluble in water. In reversed micelles it shows a preferential solvation in the oil phase. Thus the relaxation times observed in the present study may either be due to the diÜusion of the dye molecules within the interface or to the reorganization of the highly constrained bound water molecules in the interface. (c) Study of C 480 in AOT reversed micelles We have also investigated the time resolved spectra of C 480 in AOT reversed micelles and obtained the time resolved emission spectrum after subtracting the dye molecules contribution in n-heptane.The relaxation times (1.7 ns and 12 ns) and the pre-exponential factors (0.5 and 0.5, respectively) obtained by us for C(t) using a nanosecond —uorescence spectrophotometer agree with those reported by Sarkar et al.7 for C 480 at Although the 1.7 ns component is within W0\32. the —uorescence lifetime of the dye molecule, the physical meaning of the dominant 12 ns component is not trivial to comprehend.It is more likely to be representative of a process associated with the translational or rotational reorientation or diÜusion time of the reversed micelle rather than solvation dynamics of the probe dye.19 In order to understand the time resolved spectra of C 480 in AOT reversed micelles, it was necessary to reinvestigate the photophysics of this dye in this media. C 480 is a water-soluble dye22 and its —uorescence efficiency (using Quinine sulfate in 0.05 M sulfuric acid as standard) and —uorescence lifetime in water are found to be 0.62 and 5.9 ns, respectively.Both these values are close to that reported earlier.22 Structured absorption and emission spectra are obtained in n-heptane and AOT reversed micelles when very low concentration of the dye is used (concentration D5]10~6 M). The absorption and emission spectra of the dye are given in Fig. 2 and 3, respectively. The following are some of the important observations from our study.The absorption spectrum of C 480 shows a broadening at the red edge on addition of 0.1 M AOT to nheptane. In addition to spectral broadening, an isosbestic point around 390 nm»where the absorbance of the dye in n-heptane is very low»is also observed with increase in W0 . The broadening at the red edge overlaps the absorption spectrum of the dye in water. A broadening of the emission spectrum in AOT binary solution is accompanied by a loss in structure as is evident from the ratio of the —uorescence intensity of the two peaks at 390 and 410 nm.A small red-shift, in overall, of the emission spectrum is also observed. Addition of water (increase in leads to an additional broadening of the W0) Fig. 2 Absorption spectrum of C 480 in n-heptane and AOT reversed micelles. The region where the isosbestic point is observed is circled. 5032 Phys. Chem. Chem. Phys., 1999, 1, 5029»5034Fig. 3 Steady-state —uorescence emission spectrum of C 480 in nheptane and AOT reversed micelles obtained on excitation at the respective longest wavelength of maximum absorption. The region where the isoemissive point is observed is circled.—uorescence spectrum and simultaneous appearance of a broad shoulder around 480 nm. In addition, a drastic quenching of —uorescence and an isoemissive point around 480 nm is also observed. It is important to note that the isoemissive point is observed in a region where the —uorescence of the dye in n-heptane is very small and that it is observed only on addition of water.The observations made in the absorption spectra suggest that in n-heptane»AOT binary solutions, the dye C 480 is partitioned between the oil and micellar phase. An isosbestic as well as isoemissive point observed on increasing shows W0 that the distribution of the dye molecules between the micellar phase and water molecules attains an equilibrium both in ground as well as excited states.28 Fluorescence lifetimes of C 480 were also determined for diÜerent excitation and emission wavelengths and for three diÜerent values of The –ndings are given in Table 1.It is W0 . evident from the table that the emission observed at 390 and 410 nm is contributed mainly by the dye molecules in nheptane. The decay at 490 nm is mono-exponential for W0\0 Table 1 Fluorescence lifetimes and pre-exponential factors obtained of C 480 in AOT reversed micelles for diÜerent values of and at W0 diÜerent excitation and detection wavelengths.a Excitation wavelength\360 nm, Detection wavelength\390 nm W0 ] 0 4 8 32 qF (ns) 2.34 2.22 2.28 2.27 A 1 1 1 1 Excitation wavelength\360 nm, Detection wavelength\410 nm W0 ] 0 4 8 32 qF (ns) 2.41 2.26 2.33 2.38 A 1 1 1 1 Excitation wavelength\360 nm, Detection wavelength\490 nm W0 qF1 (ns) A1 qF2 (ns) A2 0 4.74 1 4 1.8 [0.39 5.2 1.0 8 1.5 [0.28 5.35 1.0 32 0.97 [0.21 5.33 1.0 Excitation wavelength\420 nm, Detection wavelength\490 nm W0 ] 0 4 8 32 qF (ns) 5.26 5.17 5.26 A 1 1 1 a Lifetime of C 480 is 2.6 ns in n-heptane and 5.9 ns in water.only. For higher values a distinct rise time is observed. The W0 bi-exponential decays with a rise time are in agreement with the isoemissive point observed at 480 nm in the steady-state —uorescence spectra. A decrease in the rise time from 1.8 ns at to 0.97 ns at is observed. In contrast, the W0\4 W0\32 decay time in the bi-exponential decays is independent (around 5.3 ns) of suggesting that this corresponds to the W0 dye molecules solubilized in the free water pool of the reversed micelle. As in the case of BC I in AOT reversed micelles,11 we have been able to determine the —uorescence lifetime of the C 480 molecules absorbing in the red-edge of the absorption spectrum.Excitation of the dye at 420 nm and observation of the —uorescence at 490 nm gave mono-exponential decays at all values of The lifetime hence obtained was also around 5.3 W0 .ns and was almost independent of This value is close to W0 . the —uorescence lifetime of the dye in water22 (see also Table 1). This proves that the dye molecules absorbing in this region are the same as those observed on excitation at 360 nm Table 1), i.e., those that are present in the free water (qF2»see pool of the reversed micelles. Sarkar et al.7 suggest that the bi-exponential decays with a nanosecond rise time are due to the ììdecrease in the energy of the guest dipole with time due to solvation so that the solvated species emitting at longer wavelength is produced on the nanosecond scale.œœ The decay time of 7 ns reported by the same authors for the dye atW0\ should be taken with caution since the —uorescence lifetime 4 of C 480 is known to gradually increase with increase in solvent polarity and reach a value of 5.9 ns in water.22 Under the circumstances the dye molecules are present in two distinct regions (interface and water pool) of the reversed micelle.It is possible then, on selective excitation of the dye and on the assumption that there is no equilibrium between the concentration of the dye in the two diÜerent microenvironments, to probe using time correlated —uorescence spectroscopy the reorganization rates of the surrounding environment that is perturbed by electronic excitation of the probe dye.28 In view of the arguments presented above, it is likely that the time-resolved spectra of C 480 in AOT reversed micelles obtained in ref. 7 and by us in the present study are representative of dye molecules in the AOT micellar interface as well as in the water pool of the reversed micelles. Thus the 1.7 ns relaxation time with a pre-exponential factor of 0.5, observed at is perhaps not a true representation of the solva- W0\32 tion dynamics of the water pool which is a solvent property. This could be associated with the diÜusion of the probe dye between the micellar interface and water pool within its excited state lifetime which is a solute property, rather than exclusively to the solvation time of the free water molecules present in the water pool of the reversed micelle.Since we have shown that it is possible to determine the —uorescence lifetime of the C 480 molecules absorbing in the red-edge of the absorption spectrum, it will be worthwhile and interesting to study the solvation dynamics of C 480 in AOT reversed micelles by exciting the dye at 420 nm and using picosecond —uorescence spectrometers. The data analysis will be more trivial due to the non-participation of the chromophores present in n-heptane or interface in the resulting time resolved emission spectra. Conclusion In conclusion, nanosecond time resolved spectra of BC I and C 480 studied in AOT reversed micelles give diÜerent results for a water insoluble dye (BC I) and a water soluble dye (C 480). The relaxation rates obtained from the study using BC I may either be due to diÜusion of the probe dye from the apolar region to the micellar interface or to the solvent relaxation times of the bound water molecules. The solvent relaxation times obtained using C 480 as a probe more likely Phys. Chem. Chem. Phys., 1999, 1, 5029»5034 5033re—ect the diÜusion times of the probe dye between the interface and free water core rather than solvent relaxation of the free water molecules present in the pool. Acknowledgements work was supported by project 2/2.1/QUI/443/94. We This are extremely thankful to Dr. Alex Siemiarczuk (PTI, Canada) for his valuable suggestions in an efficient use of the timeresolved emission spectra module in the software provided with the instrument. BBR acknowledges PRAXIS XXI for the research grant (BPD/3993/96). We are also thankful to Dr. S. Pal and Prof. S. Seshadri for the kind gift of the dyes. References 1 M. Maroncelli, J. Mol. L iq., 1993, 57, 1. 2 S. Vajda, R. Jimenez, S. J. Rosenthal, V. Fidler, G. R. Fleming and E. W. Castner, Jr., J. Chem. Soc., Faraday T rans., 1995, 91, 867. 3 G. Saielli, A. Polimeno, P. L. Nordio, P. Bartolini, M. Ricci and R. Righini, J. Chem. Soc., Faraday T rans., 1998, 94, 121. 4 (a) N. Sarkar, D. Datta, S. Das and K. Bhattacharyya, J. Phys. Chem., 1996, 100, 15483. (b) S. Matzinger, D. M. Hussey and M. D. Feyer, J. Phys. Chem. B., 1998, 102, 7216. 5 A. Datta, S. K. Pal, D. Mandal and K. Bhattacharyya, J. Phys. Chem. B., 1998, 102, 6114. 6 K. Das, N. Sarkar, S. Das, A. Datta and K. Bhattacharyya, Chem. Phys. L ett., 1996, 249, 323. 7 N. Sarkar, K. Das, A. Datta, S. Das and K. Bhattacharyya, J. Phys. Chem., 1996, 100, 10523. 8 R. E. Riter, D. M. Willard and N. E. Levinger, J. Phys. Chem., 1998, 102B, 2705. 9 J. Zhang and F. V. Bright, J. Phys. Chem., 1991, 95, 7900. 10 N. Wittouck, R. M. Negri, M. Ameloot and F. C. De Shryver, J. Amer. Chem. Soc., 1994, 116, 10601. 11 B. B. Raju and S. M. B. Costa, J. Phys. Chem. B., 1999, 103B, 4309. 12 S. Pal, PhD. Thesis, Bombay University, 1991. 13 D. V. OœConnor and D. Phillips, in T ime-correlated single photon counting. Academic Press, New York, 1984, ch. 6. 14 M. Maroncelli and G. R. Fleming, J. Chem. Phys., 1987, 86, 6221. 15 B. B. Raju, J. Phys. Chem., 1997, 101A, 981. 16 B. B. Raju and S. M. B. Costa, Phys. Chem. Chem. Phys., 1999, 1, 3539. 17 B. B. Raju and B. Eliason, J. Photochem. Photobiol., A: Chem., 1988, 116, 135. 18 H. Shirota and K. Horie, J. Phys. Chem., 1999, 103B, 1437. 19 P. E. Zinsli, J. Phys. Chem., 1975, 83, 3223. 20 The authors in ref. 8 report that the mole fraction of intrinsic water to AOT as determined by Karl-Fischer titration, for (W0), the AOT sample used to be 1. The results reported for in W0\0 ref. 8 then should actually correspond to the results at W0\1 and that at to the results obtained at W0\2.8 W0\3.8. 21 W. Jarzeba, G. C. Walker, A. E. Johnson an P. F. Barbara, Chem. Phys., 1991, 152, 57. 22 G. Jones, II ; C. Y. Choi, W. R. Jackson and W. R. Bergmark, J. Phys. Chem., 1974, 89, 294. 23 H. Cho, M. Chung, J. Lee, T. Nguyn, S. Singh, M. Vedamuthu, S. Yao, S.-B. Zhu and G. W. Robinson, J. Phys. Chem., 1995, 99, 7806. 24 M. Bellete� te, M. Lachaoelle and G. Durocher, J. Phys. Chem., 1990, 94, 5337. 25 H. L. Pasquier, R. B. Pansu, J.-P. Chauvet, P. Pernot, A. Collet and J. Faure, L angmuir, 1997, 13, 1907. 26 W. Rettig and M. Zander, Ber. Bunsen-Ges. Phys. Chem., 1987, 87, 1143. 27 N. Mataga, H. Yao, T. Okada and W. Rettig, J. Phys. Chem., 1989, 93, 3383. 28 S. M. B. Costa, M. M. Velaç zquez, N. Tamai and I. Yamazaki, J. L umin., 1991, 48 & 49, 341 and references therein. Paper 9/06191F 5034 Phys. Chem. Chem. Phys., 1999, 1, 5029»50