Langmuir 1989,5,942-947
942
Fluorescence Quenching by Oxygen in Reverse Micellar Solutions M. SBez, E. A. Abuin, and E. A. Lissi* Departamento de Qulmica, Facultad de Ciencia, Universidad de Santiago de Chile, Casilla 5659, Correo 2, Santiago, Chile Received July 25, 1988. I n Final Form: March 1, 1989
The quenching of the fluorescence of micelle-solubilized probe molecules by oxygen has been studied in reverse micellar solutions. Positive ((11-(1-pyrenyl)undecyl)trimethylammonium,(4-(l-pyrenyl)butyl)trimethylammonium, (( 1-pyrenyl)methyl)trimethylammonium,and tris(2,2'-bipyridine)ruthenium(II)) and negative probes (pyrenesulfonate, pyrenetetrasulfonate, and tris(4,4'-dicarboxylate-2,2'-bipyridine)ruthenium(I1)) incorporated either to anionic (sodium bis(2-ethylhexyl)sulfosuccinate (AOT)/water/ chloroform, AOT/water/isooctane, and (AOT)/water/dodecane) or cationic (cetyltrimethylammonium chloride/water/chloroform) reverse micelles were considered. The study was performed covering a wide range of water to surfactant mole ratios. Probes which bear a single charge and can be solubilized at the micellar surface, intercalated between the surfactant heads, are quenched by oxygen with rates dependent on the oxygen solubility and mobility in the dispersion media. The quenching rate is also dependent on the distance between the chromophore and the group bearing the charge, but it is not very dependent on the sign of the probe charge. Probes which due to their large and delocalized charge cannot be incorporated deeply into the micellar interface are quenched with rates considerably lower than those found for probes anchored at the micellar surface. This effect is interpreted in terms of a restricted access of oxygen toward the water pools.
Introduction There exists a wealth of information on the interaction of ground-state oxygen with excited molecules incorporated to or adsorbed at the surface of hydrophobic assemblies of surfactants in aqueous solutions.l+ Since the deactivation of the excited molecule is diffusion controlled, the rate of the process will be determined by the local concentration of oxygen a t the probe microenvironment and by the mobility of the oxygen m ~ l e c u l e . The ~ , ~ quenching rate constant will then be sensitive to the characteristics of the interface and to the relative location of the probe. Similarly, it also will be depend. n t upon the relative solubility of oxygen in the disperse microphase and the surrounding medium. The incorporation to the microphase of an oxygen molecule reaching the interface from the dispersion medium will be an efficient process when the intracompartment solubility is larger than that in the dispersion medium (as in micelles or but must be inefficient when the intracompartment solubility is smaller (as in reverse micelles or water-in-oil microemulsionsg). In this context, it is interesting to compare quenching rates in normal and reverse micellar solutions. Nevertheless, there are very few studies directed to examining the effect of reverse micelles upon the interaction rate of micelle-solubilized excited probes with oxygenlO~*l (1) Turro, N. J.; Aikawa, M.; Yekta, A. Chem. Phys. Lett. 1979, 64, 473. (2) (a) Geiger, M. W.; Turro, N. J. Photochem. Photobiol. 1975, 22, 273. (b) Gorman, A. A,; Lowering, G.; Rodgers, M. A. J. Photochem. Photobiol. 1976, 23, 399. (c) Lakowicz, J. R.; Weber, G. Biochemistry 1973, 12, 4161. (3) Vaughan, M. W.; Weber, G. Biochemistry 1970, 9, 464. (4) Lissi, E. A,; Dattoli, A. M.; Abuin, E. B. Bol. SOC.Chil. Quim. 1985, 30, 37. (5) Calhoun, D. B.; Vanderkoii, J. M.; Woodrow, I. G. W.; Englander, S. W. Biochemistry 1983, 22, 1526. (6) Rubio, M. A.; Araya, L.; Abuin, E. B.; Lissi, E. A. An. Asoc. Quim. Arg. 1985, 73, 301. (7) Geiger, M. W.; Turro, N. J. Photochem. Photobiol. 1977, 26, 221. (8) (a) Matheson, I. B. C.; King, A. D. J. Colloid Interface Sci. 1967, 66, 768. (b) Matheson, I. R. C.; Massoudi, R. J . A m . Chem. SOC.1980, 102, 1942. (9) Lee, P. C.; Rodgers, M A . I .I. Pliys. Chpm. 1983, 87, 4894.
0743-7463/89/2405-0942$01.50/0
and only one on the effect of the probe location upon the quenching rate." In this last work, assuming that the alkyl chains of pyrenyl(CH2),C00- probes are completely extended, it is concluded that an oxygen concentration gradient exists in the reverse micelles formed by calcium alkyl benzenesulfonates containing inorganic materials and that the gradient is qualitatively dependent on the amount of inorganic material solubilized in the polar core. The effect of the core size was also evaluated by Wong et al.'O for the quenching of the fluorescence of pyrenesulfonate in the AOT/water/ heptane reverse micellar system. A small decrease in the quenching constant was observed when increasing the water to surfactant mole ratio ( R = [H,O]/[AOT]), and it was concluded that oxygen can diffuse with great east toward the site where the fluorophore is solubilized, even at low R values, where the microviscosity is very high. The quenching of the fluorescence of several pyrene derivatives by oxygen was also measured by Atik and Thomas in cetyltrimethylammonium bromide/ water/ hexanol/dodecane water-in-oil microemulsions.'* This work was carried out at a single microemulsion composition ([water]/ [CTAB] = 20.3; [hexanol]/[CTAB] = 2.2), and it was observed that the pseudounimolecular quenching rate constant decreased from 4.2 X lo7 s-l for pyrene to 1.2 X lo7 s-l for pyrenesulfonate. In the present work, we report results obtained in a study of the effect of reverse micelles upon the fluorescence quenching of probe molecules by oxygen. The systems considered involve cationic (( 11-(1-pyreny1)undecyl)trimethylammonium, (4-(l-pyrenyl)butyl)trimethylammonium, and ruthenium bipyridyl) and anionic (pyrenesulfonate, tris(4,4'-dicarboxyl-2,2'-bipyridine)ruthenium(II), and pyrenetetrasulfonate) probes solubilized in anionic (sodium bis(2-ethylhexyl)sulfosuccinate/water/ organic solvent) and cationic (cetyltrimethylammonium chloride/water/chloroform) reverse micellar solutions. (10) Wong, M.; Thomas, J. K.; Gratzel, M. J . Am. Chem. Sot. 1976, 98, 2391. (11)Tzi-Chi, J.; Kreuz, K. L. J. Colloid Interface Sci. 1984, 102, 308. 112) Atik, S. S.; Thomas, J. K. J. Am. Chem. SOC.1981, 103, 4367.
0 1989 American Chemical Society
Langmuir, Vol. 5, No. 4, 1989 943
Fluorescence Quenching by Oxygen
Table I. Quenching of the Fluorescence of Pyrene Derivatives by Oxygen in AOT/Hydrocarbon/Water rN," ns
probe methylpyrene
solvent isooctane dodecane
PUTMA
AOT
R 1
water AOT
PMTMA
PS
1 5 10 20 30
water AOT
PTS
1 5 10 20 30
water AOT
1 5 10 20 30
water AOT
kQb
dodecane
isooctane
250
10 20 30 PBTMA
isooctane
1 6 25
dodecane
10.0
240 230 227 220
248 232 217 215 135 232 222 220 216 210 60 147 123 122 112 112 64 139 109 104 102 98 11.7 9.3 9.3 10.9
3.7 2.6 2.4
6.2 5.8 5.1 5.4 0.24 4.5 4.0 3.8 3.6 3.6 0.24 3.8
2.4 2.4 2.1 2.0 1.9 1.8 1.7 1.8 1.7
3.4 3.2 3.0 2.8 0.22 3.4 2.8 2.7 2.3 2.2 0.20 50.1 50.15 50.15
143 112 105 101
1.9 1.7 1.5 1.4
'Estimated error (PUTMA, PBTMA, PMTMA, and PSA) f 3 % ; PTSA = 10.5 na. *In lo' s-l. Error estimated from lifetime uncertainties: 5%.
Experimental Section Sodium bis(2-ethylhexyl)sulfosuccinate (AOT; Aldrich) was purified as previously de~cribed.'~Cetyltrimethylammonium chloride (CTAC) was purified by repeated recrystallizations from acetone-methanol mixtures. (11-(1-Pyreny1)undecyl)trimethylammonium iodide (PUTMA), ((1-pyreny1)methyl)trimethylammonium iodide (PMTMA), (4-(l-pyrenyi)butyl)trimethylammonium bromide (PBTMA), 1-pyrenesulfonic acid, sodium salt (PS), and 1,3,6,&pyrenetetrasulfonic acid, sodium salt (PTS) from Molecular Probes and ruthenium bipyridyl chloride 6H20,(Ru(bpy)32+(G.Frederich Smich Chem. Co.) were employed as received. Their fluorescence spectra and fluorescence lifetimes were in agreement with reported values. Tris(4,4'-dicarboxyl2,2'-bipyridine)ruthenium(II), (RuL:-) was a gift of Dr. Joao Bonilha (University of Sao Paulo, Riberao Preto, Brasil) and was synthesized according to the procedure described by Foreman et al.I4 Chloroform (B&J, high purity) was employed after washing with water, drying over calcium chloride, and distillation. For most probes employed, fluorescence lifetimes were measured (in nitrogen-purged and in air- or oxygen-saturated solutions) by following the luminescence decay after excitation with the pulse of a fast nitrogen laser (Nitronite LN 100). Fluorescence intensities were measured from 20 ns after excitation (10)down to nearly 0.1510. Plots of ln (p/n vs time were lineal (correlation coefficient I 0,999with negligible intercept). The reproducibility of the lifetimes evaluated from independent measurements was 3 % . PTS lifetimes were measured by using a Gregg phase shift spectrofluorimeter. Phase and modulation lifetimes gave a single value when the excitation frequency was changed from 6 to 30 MHz. All experiments were performed at room temperature in solutions at 0.1 M surfactant concentration. Surfactants to probe ratios were always >lo3 to avoid multiple probe occupancy. Oxygen and nitrogen purging of the solutions was carried out at very small rates to minimize solvent and/or water evaporation. In order to avoid water evaporation that could modify the water/surfactant ratio, some experiments were carried out by (13) Maitra, A. N.; Eicke, H.F. J . Phys. Chem. 1981, 85, 2687. (14) Foreman,T. K.; Sobol, W. M.; Whitten, D. G. J.Am. Chem. SOC. 1981, 103,5333.
Table 11. Quenching of the Fluorescence of Pyrene Derivatives by Oxygen in AOT/Chloroform/Water probe solvent R iN, ns r4: IO7 s-l PUTMA chloroform 57 3.3 AOT
PBTMA
chloroform AOT
PMTMA
1 3 5 7
chloroform AOT
PS
1 3 5 7
1 3 5 7
chloroform AOT
1 3 5 7
54 55 56 56 65 73 75 77 80 31.5 62 60 60 57 77 74 68 66 66
3.1 3.1 3.1 3.2 3.1 2.7 2.7 2.8 2.7 3.4 2.4 2.2 2.0 1.9 3.1 1.9 1.7 1.7 1.8
'Estimated error = f O . l x IO7 s-l. adding water to prepurged surfactant solutions. Similar results were obtained.
Results and Discussion Fluorescence intensities, either i n the absence (nitrogen purged) or the presence of air, can be adequately fitted to a monoexponential decay. Monoexponential decays in the absence of quenching imply that the excited molecules constitute a single and homogeneous population, as expected for probes which are totally incorporated to the reverse micelles. Monoexponential decays i n the presence of quenching are a consequence of the fact that fluorescence intensities are measured starting 20 ns after excitation, which avoids a n y multiexponential decay due to
944 Langmuir, Vol. 5, No. 4, 1989
Sciez et al.
Table 111. Quenching of the Fluorescence of Pyrene
Table V. Experimental Data for RuL3"
Derivatives by Oxygen in CTAC/Water/Chloroform probe solvent R TN," ns kq?b lo7 s-l methylpyrene PUTMA
PBTMA
chloroform chloroform CTAC
1 3 5 10 15 20
chloroform CTAC
1 3
5 7 10 15 20 PMTMA
PS
PTS
chloroform CTAC
1 1.5 3 5 10 15 20
chloroform CTAC
CTAC
1 3 5 7 10 15 20 1 5
53 54 57 (58) (59) 60 (60) 62 65 66 65 71 (67) 82 87 (77) (80) 94 98 102 32 44 (38) (39) 56 60 (56) 67 68 71 77 87 (95) 90 (90) 91 (97) (86) 93 92 92 (4.4) (4.4)
solvent ethano1:water (91) water (pH 11) AOT/isooctane
4.1 3.1 3.2 (3.0) (2.9) 3.0 (2.6) 2.7 2.6 3.1 3.3 (2.6) 2.6 2.2 (1.6) (1.4) 1.8 1.8 1.4 3.4 3.0 (2.0) 2.2 1.9 (1.4) 1.5 1.5 1.4 3.1 2.3 (2.0) 2.2 (1.7) 2.0 (1.5) (1.4) 1.7 1.4 1.35 (1.2)C (1.2)
" Values between parentheses were obtained in the absence of ethanol. bEstimated error: fO.l X lo7 s-l. cError: f 0 . 2 X s-l. Table IV. Experimental Data for Ru(bpy),2+ solvent (viscosity) R (AJmmr i N . ns ko, lo6 s-' 1250 4.6 acetone (0.30) 606 960 4.4 acetonitrile (0.34) 740 3.1 methanol (0.55) 600 600 0.77 water (1.0) 800 2.4 2-propanol (2.9) 1.9 610 825 1-heptanol (7) 1 (616)" (1020)" (240 CTAC/chloroform 2 1100 2.0 (2.3)" 3 (618)" (1000)" (990)" (2.1)" 5 (619)" 900 2.0 6 624 8 830 2.1 780 1.5 10 621 720 1.0 15 615 700 0.9 20 607 700 0.9 25 1 c 648 (675)b 1.8 (l.O)b AOT /isooctane 5 612 (620)b 2.5 (l.8)b 610 (620)* 2.5 (2.0)b 10 605 (620)b 2.5 (2.1)b 15 610 (620)b 2.5 (1.9)b 25 610 2.5 30
khimd
1.92 2.2 1.5 3 1.1 1.2
" Obtained in absence of ethanol.
bObtained in AOT/dodecane. was nearly 618 nm for all R values. dBimolecular rate where [O,]is the constant in los M-' s-l defined by kbim= ke/[Oz], oxygen concentration in an air-saturated solution.
c(A.,m)-
transient terms, the low oxygen concentration (see below), and its high mobility. These last two conditions leads to a completely dynamic quenching, affording monoexponential decays and a behavior that can be discussed in terms of the Stern-Volmer law. The results obtained are given in Tables I-V. In these tables, kQ,the pseudounimolecular rate constant for the
CTAC/chloroform"
R 13 16 18 21 23 30 35 45 2.5 5 7.5 10 15
(A),
TN,
607 618 619
619 620 618 620 620
ns
707 700 720 683 672 650 650 630 625 1050 920 880 843 825
kq, lo6 s-'
0.62 0.39 0.41 0.41 0.38 0.39 0.39 0.50 0.53 0.34 0.33 0.38
" Chloroform containing 1% (v:v) ethanol was employed. The probe was insoluble in micelles prepared in the absence of ethanol. deactivation of the excited probes by oxygen in air-saturated solutions, is defined by or
where T ~ T N~, and , T~~ are the fluorescence lifetimes measured in air-, nitrogen-, and oxygen-saturated solutions, respectively. For the ruthenium complexes, the wavelengths at the maximum of the emission spectra, (A,), are also included in Tables IV and V. Pyrene Derivatives in AOT/Water/Isooctane. The results obtained in this system are shown in Table I. The lifetimes of PBTMA, PMTMA, and PS in the micelles are considerably larger than in water, an effect that is readily explainable in terms of the quenching of fluorescence of pyrene and pyrene derivatives by polar solvent^.^ For all the compounds considered but PTS, the lifetimes under nitrogen decreases when R = [water]/ [surfactant] (mole ratio) increases, indicating that the probes experience a more polar environment. It is remarkable that even the lifetime of PUTMA decreases from 248 ns ( R = 1)to 215 ns (R = 30),showing that even in this compound the pyrenyl group senses the polar interface. The pseudounimolecular deactivation rate constants in the micellar solutions are, for all the compounds but PTS, intermediate between those obtained in isooctane (methylpyrene as probe) and water. The range of R values considered comprises the region where a clear "pool" cannot be defined ( R I8) and the region of large R values, where a distinct water rich core can be ~0nsidered.l~ The mean occupation number of the water pools by oxygen is very low. A t the larger R value employed ( R = 30),a water pool contains -1.8 X lo4water molecules. In air-equilibrated solutions, the mole fraction of oxygen in If the oxygen solubility in the water water is 0.5 X pool of the micelle is considered to be equal to that in bulk water, the data given above imply that only one-tenth of the water pools contain a dissolved oxygen molecule. Under these conditions, kQ will be determined by the oxygen diffusion rate in the dispersion media and by the probability (a)that an oxygen-micelle encounter leads to the quenching. If this probability is high, kQ can be expressed by (15) (a) Bardez, E.; Gouguillon, B. T.; Keh, E.; Valeur, B. J. Phys. Chen. 1984,88, 1909. (b) Maitra, A. J. J. Phys. Chem. 1984,88, 5122. (16) Battino, R.; Rettich, T. R.; Tominaga, T. J . Phys. Chem. Ref. Data 1983, 12, 163.
Langmuir, Vol. 5, No. 4, 1989 945
Fluorescence Quenching by Oxygen kQ
=
(kdif)isooctanelY[021isooctane
(3)
The value of a can be considerably smaller than 1due to either the protection established by the interface as a consequence of its high viscosity (“kinetic” barrier) or low oxygen solubility (“thermodynamic” barrier). Even if the interface does not impose an extra barrier, the rate of entrance (i.e., the probability that an oxygen molecule reaching the micelle surface goes to the micelle interior) will be determined by the relative solubility in the micelle interior and in the surrounding medium.l’ In reverse micelles, the lower solubility of oxygen in the aqueous microphase would lead to “jumping out” rate constants faster than the rates of “jumping in”. Under these conditions, k- (the exit rate) would be faster than k, (the pseudo unimolecular intramicellar quenching constant). In this case, kQ would be determined by the equilibrium intramicellar oxygen concentration and will be given by kQ
=
(kdif)micelle[021micelle
(4)
where is the diffusion-controlled bimolecular rate constant in the microphase. Equations 3 and 4 represent extreme situations and will be employed as a reference in discussing the data obtained. PBTMA, PMTMA, and PS show similar values of kQ, which slightly decreases when R increases. All the probes are then similarly exposed to the incoming oxygen, indicating that, irrespective of the charge and the value of R, the pyrenyl group must be located in the nonpolar region (actually the nearly 2 factor of protection can be accounted for by assuming a reduced solid angle from where the oxygen can diffuse toward the probe18). The small decrease of kQ observed when R increases could indicate a moderate displacement of both positive (PBTMA and PMTMA) and negative (PS) probes toward a more polar environment. This is also supported by the decrease of rN obtained for all these probes when R increases. PUTMA shows values of kQ larger than those for the other pyrene derivatives. This is the result expected if consideration is given to the larger number of spacers (CH2 groups) between the ionic head and the fluorophore. Nevertheless, it is remarkable that even for this compound kQ and T N are considerably smaller than those measured for methylpyrene in isooctane and that they also show a small decrease when R increases. This effect is consistent with the known tendency of the pyrene moiety to be solubilized at micelle-water interfaces and is in agreement with data obtained by Whitten et al.19 from the effect of R upon the luminescence spectra of pyrene in the same reverse micellar system. The relevance of oxygen diffusion in the dispersion medium is stressed by the data obtained by employing dodecane as a nonpolar component. The results given in Table I show that the lifetimes under nitrogen obtained by employing dodecane are similar to those measured when isooctane is the dispersion medium, but kQ values are considerably smaller. In bulk solvents, the value of (kQ)isooctane/(k$)do&me equal to 2.7 Can be accounted for in terms of the difference in the solvent viscosities. The relative values of (kQ)isooctane/(kQ)ddecane are smaller in the micellar solutions than in the bulk solvents and depend both on the probe considered and the value of R. In particular, the data given in Table I show that PUTMA presents the larger (kQ)isooctane/ (kQ)d&cane values, as ex(17)Almgren, M.; Grieser, F.; Thomas, J. K. J.Am. Chem. SOC.1979, 101. 279. -, (18)Thomas, J. K. Chem. Reu. 1980,80, 283. (19)Backer, C. A.; Whitten, D. G. J. Phys. Chem. 1987, 91, 865.
pected due to its deeper penetration into the hydrocarbon pseudophase. Equation 3 predicts that, for processes near the diffusion-controlled limit, kQ values will be determined by the oxygen solubility in the surrounding media. On the other hand, for slower processes that are not controlled by the diffusion in the solvent and under conditions such that eq 4 applies, the observed rate constant should become independent of the oxygen solubility and mobility in the dispersion media. The results given in Table I are in agreement with these considerations. If the observed kQ values in air-equilibrated solutions are divided by the oxygen solubilities in isooctane (0.0032 M)16 and dodecane (0.0019 M),16the bimolecular rate constants obtained for PUTMA are still larger in isooctane than in dodecane, a result predicted by eq 3. On the other hand, for slower reactions (i.e., those of PS), diffusion in the bulk is less determining, and the rates observed are much less dependent upon the solvent viscosity (as predicted by eq 4). PTS can be expected, when R >> 8, to be incorporated to the water pools.20p21 Due to the short lifetime of this probe, reliable values of It, are difficult to obtain under the experimental conditions employed. However, it is remarkable that, over all of the [H20]/[AOT] range considered, the values of k are considerably smaller than and even smal er than (kQ)water,indicating that the interface establishes a strong protection barrier. This result is particularly remarkable a t low values of R, where a distinct water pool is not present. The difference between the kQ value observed for PTS with regard to those obtained for the other pyrene derivatives, which are intercalated between the surfactant heads at the micellar interface, implies that even a t low values of R PTS must be located at the center of the micellar aggregates21*22 and that oxygen penetration into this region is very slow. The fact that at large R values the quenching in the water pools is slower than in bulk water can be explained in terms of a slow oxygen diffusion through the interface and/or low values of (kdif)miceue or [O2lmiCeh.Nevertheless, since singlet oxygen reactions, which are not diffusion controlled, take place in the AOT water pools at high R values with similar rates than in bulk water?23 it can be argued that the oxygen solubility in the micelle is not significantly lower than in bulk water. The low 12 observed must then reflect a kinetic limitation impose! by the interface and/or the higher viscosity of the pool.21 Pyrene Derivatives in AOT/ Water/Chloroform. The results obtained when employing chloroform as a dispersion media show noticeable differences with regard to those obtained in the micellar solutions in isooctane. The lifetimes of the probes are considerably shorter due to quenching by c h l o r o f ~ r m . The ~ ~ incorporation of the probes into the micelles (or to the region where a significant concentration of surfactant tails exists) increases the probe’s lifetimes. This effect is more noticeable for PMTMA, but it is also significant for PBTMA. Regarding the oxygen quenching, it can be seen that the values of (kQ)hlk/ ( k ~ are) smaller ~ ~ than ~ those obtained by employing isooctane (a solvent of similar viscosity) as the dispersion medium. This effect can be explained in terms of an easier penetration of oxygen through the interface, likely due to a more favorable oxygen partitioning between the micelle and the surrounding solvent. The lower solu-
B
(20) Verbreck, A.; DeSchryver, F. C. Langmuir 1987, 3, 494. (21) Keh, E.; Valeur, B. J. Colloid Interface Sci. 1981, 79, 465. (22) Kondo, H.; Miwa, I.; Sunamoto, J. J . Phys. Chem. 1982,86,4862. (23) Rubio, M. A,; Lissi, E. A. J. Colloid Interface Sci., in press. (24) Encinas, M. V.; Rubio, M. A.; Lissi, E. A. Photochem. Photobiol. 1983, 37, 125.
946 Langmuir, Vol. 5, No. 4, 1989
Sdez et al.
bility of oxygen in the chloroform (relative to that in isooctane) is stressed by the smaller value of kQ observed in bulk chloroform relative to that found in isooctane. Pyrene Derivatives i n CTAC/ Water/Chloroform. CTAC forms, in chloroform, reverse micelles which can Cationic surfactant take only a limited amount of water (up to R = 7 ) . The reverse micelles produced are small, and their properties will be strongly sensitive to the presence of the water More stable micelles, which are able to take a large amount of water and which probably are considerably larger,12are obtained if 1%of ethanol (that would act as cosurfactant) Anionic surfactant is added to the chloroform. The results obtained in these systems are collected in Table 111. The probe's lifetimes in the CTAC/water/chloroform systems show as for the AOT/ water/ chloroform systems (see Table 11) that CTAC addition increases the probe lifetime. Furthermore, for cationic probes the effect increases when R increases. The data of Table 111also show that this protection is significantly larger in the presence of ethanol, particularly a t low R values. With regard to RU L - ~ 3 quenching by oxygen, all probes show a decrease in kQ when R increases. Furthermore, it is interesting to note that the relative values of (kQ)mAC/(kQ)Am are, particularly when R is low, larger for PS than for PTMA, a result that Cationic surfactant can be explained in terms of a more internal location of the co-ions. The data of Table 111 show that alcohol addition originates a noticeable increase in the quenching rate E * water constant. Since the effect of the alcohol upon rNis not compatible with a more external (Le., toward the chloro. , form) location of the probes, the results obtained point to Anionic surfactant a more favorable penetration of oxygen into the ethanol modified micelle's interface. Quenching of the Emission of Ruthenium Comwater > H plexes. Transition-metal complexes, in particular ruthenium(I1) complexes, have been extensively employed as probes in the study of micelles26 and monolayer^.^' Relevant to the present discussion is the fact that their Figure 1. Relative location of Ru(I1) derivatives at low (A, C, photochemistry is highly solvent d e ~ e n d e n tand ~ ~ that !~~ E, G) and large (B, D, F, H) water/AOT values in reverse micelles quenching by oxygen, a t least in water, is considerably formed by cationic and anionic surfactants. slower than for the pyrene derivatives. For Ru(bpy):+ in the efficiency of the excited-state deactivation by oxygen. water at 25 "C, Demas et aL30have reported a bimolecular Ru(bpy)QS+ i n Anionic a n d Cationic Reverse Miquenching rate constant of 1.8 x lo9 M-' s-l , a value celles. Due to its positive charge, it can be expected that slightly smaller than the theoretical upper limit for oxygen Ru(bpy),2+,in negative micelles, is going to be incorporated quenching of triplet states by an energy-transfer mechaat the micellar surface by intercalation between the surnism. Smaller quenching rates obtained for other comfactant heads, for all R values. If the position of the probe plexes were explained in terms of rather tight solvation is assessed from its T N and ,,,),,A( the data of Table IV spheres, which retard penetration of oxygen to an effective permit one to conclude that from R I 5 to 30 the probe quenching distance. microenvironment remains nearly constant and that it is Table IV shows the data obtained for R ~ ( b p y ) , ~in+ similar in AOT/water/isooctane and AOT/water/dodedifferent solvents and in anionic and cationic reverse cane micellar systems. Similarly, kQ values remain conmicelles. These data show that, in homogeneous solvent, both T N and (A,) are highly solvent d e ~ e n d e n t . ~ ~ , stants ~ ~ over all the range covered for R. The value obtained in AOT/water/isooctane (kQ = 2.5 (hO.1) X lo6 s-l) is Furthermore, the results shown in this table also indicate larger than that measured in AOT/water/dodecane (kQ that the bimolecular quenching rate constants for deac= 2.0 (A0.l) X lo6 s-l), but the difference is considerably tivation of excited R ~ ( b p y ) by ~ ~oxygen + are not simply smaller than that found for the pyrene derivatives, indirelated to the solvent viscosity, indicating that other factors cating that the mobility of oxygen in the dispersion media (such as changes in solvation and/or changes in the nature is not the mean factor involved in determining the of the excited state) play a significant role in determining quenching rate of Ru(bpy)?+ emission. For R = 1, both rN and kg values are significantly dif(25) Fendler, J. H.; Fendler, E. J.; Medray, R. T.;El Seoud, 0. A. J . ferent than those obtained at larger R values. The difChem. SOC.,Faraday Trans. I 1973, 69, 280. ference could be related with ionic strength effects favoring (26) (a) Hauenstein, B. L., Jr.; Dressick, W. J.; Buell, S. L.; Demas, J. ion pairing under these condition^.^^ Nevertheless, we N.; De Graff, B. A. J. Am. Chem. SOC.1983,105,4251. (b) Hauenstein, B. L., Jr.; Dressick, W. J.; Gilbert, T. B.; Demas, J. N.; De Graff, B. A. found no effect of salt addition (3 M NaC1) upon either J. Phys. Chem. 1984,88, 1902. the lifetime or the quenching rate constant by oxygen in (27) Sprintschnik, G.; Sprintschnik, H. W.; Kirsh, P.K.; Whitten, D. water. Furthermore, i t has been shown that R ~ ( b p y ) , ~ + G. J . Am. Chem. SOC.1977, 99, 4947. (28) Caspar, J. V.; Meyer, T. J. J . Am. Chem. SOC.1983, 105, 5583. (29) Nakamaru, K. Bull. Chem. SOC.Jpn. 1982,55, 1639. (30) Demas, J. N.; Harris, E. W.; McBride. R. P. J. Am. Chem. SOC. 1977, 99, 3547.
(31) We thank one of the reviewers for suggesting we address this point.
Fluorescence Quenching by Oxygen
emission is not affected by ionic strength effects (9 M LiC1) in aqueous solution.32 The nearly 2 factor of decrease in the quenching rate by oxygen can then be considered as indicative of a much reduced solubility and/or oxygen mobility in the surroundings of the probe under these conditions of low surfactant head hydration. The behavior of Ru(bpy)32+in the cationic micelles is clearly different (see Table IV). Both , ,),A,( and T N steadily decrease when increasing R, approaching values very similar to those measured in bulk water, pointing to a displacement of the probe toward the micelle interior. Similarly, kQ decreases toward a value which also is close to that measured in water. This is the behavior expected from eq 4 if the probe is displaced toward the center of the pool and the quenching is not taking place at every probe-oxygen encounter. RuL,*- in Cationic and Anionic Reverse Micelles. The solubility of RuL,4- in the reverse micelles is considerably smaller than that of the unsubstituted Ru(bpy)32+. In AOTIwaterlisooctane, it can only be dissolved when R 1 3. In CTAC/water/chloroform, clear solutions were obtained only in the presence of 1% ethanol. (Aem)- and T N values do not show a clear trend with R although a tendency to a decrease in T N when increasing R is observed in both micelles (see Table V). The values of kQ show a behavior markedly different than that for the quenching of Ru(bpy)32+. When R is small, kQ values are, in both micellar systems, smaller than those for Ru(bpy)?, indicating that, as a consequence of the charge of the ligands, R u L ~ is ~ -located toward the micelle interior, irrespective of the surfactant charge. In the micelle formed by the cationic surfactant, the value of kQ remains nearly constant independently of R, a result that can be interpreted in terms of a nearly constant location (the inner side of the interface) over all the R range. On the other hand, in the micelle formed by the anionic (32) Allsopp, S. R.; Cox, A.; Kemp, T. J.; Reed, W. J. J. Chem. SOC., Faraday Trans. 1 1978, 74, 1275.
Langmuir, Vol. 5, No. 4 , 1989 947
surfactant kQ increases when R increases, approaching, as in the case of R ~ ( b p y ) , ~ a+ ,value very similar to that measured in bulk water. The different effect observed for both ruthenium complexes when they are co-ions or counterions a t low and high R values can be naively interpreted in terms of locations, such as those depicted in Figure 1. Locations of PS as depicted in Figure 1A and of PTS as depicted in Figure 1E can also explain the large difference in the kQ values found for these probes in AOT micelles a t low R values.
Conclusions The interaction rate between excited states of probe molecules and oxygen in reverse micellar solutions is dependent on the location of the chromophore and the micellar size. Probes anchored a t the micellar surface and exposing the chromophore toward the organic pseudophase interact with oxygen with a rate that is nearly independent of the charge of the probe and is mainly determined by the oxygen solubility and mobility in the dispersion media, but it is still dependent on the micellar size as a consequence of probe relocations. Probes which solubilize with exposition of the chromophore toward the micellar interior interact with oxygen with smaller rates, emphasizing the importance of a micelle restricting access of oxygen to the water pools. This effect has not been observed in solutions of normal micelles in water and can be related to the lower solubility of oxygen in the aqueous interior of reverse micelles with regard to that in the surrounding medium. Acknowledgment. We are grateful to DICYT (Universidad de Santiago de Chile) and FONDECYT (Grant 507187 and 4012/86) for financial support. E. B. Abuin is grateful to the John Simon Guggenheim Memorial Foundation for a fellowship. Registry No. AOT, 577-11-7; CTAC, 112-02-7; PUTMA, 103692-03-1; PBTMA, 81341-11-9; PMTMA, 72185-47-8; PS, 59323-54-5;PTS, 59572-10-0;02,7782-44-7; RU(bpY)?+, 1515862-0; RuL$-, 78338-26-8; CHCl,, 67-66-3; isooctane, 540-84-1; dodecane, 112-40-3.