J . Chem. Soc., Faraday Trans. 1, 1982, 78, 2167-2181 Quenching of Fluorescence from Ru(digy)i+ and Ru(dipy),(CN), in Solutions of Sodium Dodecyl Sulphate B Y STEPHEN J. ATHERTON, JOHN H. BAXENDALE* AND BRIDGID M. HOEY Department of Chemistry, University of Manchester, Manchester M 13 9PL Received 7th September, 198 1 The kinetics of the fluorescence quenching of Ru(dipy)i+ and Ru(dipy),(CN), by Cu2+ and MVZ+ are reported. At SDS concentrations above the c.m.c. the kinetics of Ru(dipy)i+ quenching by both quenchers are simple first order with constants which increase linearly with quencher concentration and which decrease with increasing SDS concentration. The results quantitatively agree with those predicted by the model in which the emitter is completely micellised and the quenching is slow enough to allow equilibration of the quencher between solution and micelles.With Ru(dipy),(CN), above the c.m.c., the quenching by MV2+ has fast and slow components, the latter being that of the natural emission decay. The results are shown to be consistent with almost complete micellisation of the quencher and a high micellar quenching rate constant. With Cu2+ the micellar quenching is too fast to follow so that ‘quasi-static quenching’ is seen, i.e. there is a decrease in the initial emission intensity followed by the slower unquenched decay. However, analysis shows that specific interaction of Cu2+ with the emitter on the micelle must be assumed to account for the abnormally high efficiency of Cu2+. Enhanced quenching of Ru(dipy)i+ emission by Cu2+ and MV2+ is also observed below the c.m.c.This is explained in terms of the formation of clusters of the emitter and SDS with which the quenchers associate. With Ru(dipy),(CN), only Cu2+ gives enhanced quenching below the c.m.c. It is suggested that Cu2+ complexes with SDS to form a micelle-like species with which the emitter can associate. The kinetics of fluorescence quenching in micellar systems are determined by a number of factors such as equilibrium constants for the distribution of emitter and quencher between micellar and aqueous phases, the rates of transfer of emitter and quencher to and from the micellar and aqueous phases, the rate of quenching in the micellar phase and the natural lifetime of the emitter. For the simpler case where the emitter is completely micellised and the quencher only partially so, these processes can be represented: P* -+ P P*+nQ, + P k+ Qa+M*Qm k- ko nkq, quenching with n quenchers in the same micelle K , k,, k-, exchange of Q between water and the micelles where P is the emitter, Q, and Qm are the quencher in aqueous and micellar phase, M is the micelle and K = k + / k - = [Qm]/[Q,][M], where [MI is the total micelle concentration.Infeltal has derived an equation for quenching kinetics which takes account of the above processes and which has been amended recently by Dederen et al.2 to include an additional process, viz. the transfer of quencher directly from one micelle to another: WnQ) + MbQ) -+ W z Q ) + M(P + 1Q) k , 21672168 QUENCHING OF FLUORESCENCE FROM Ru(dipy)t+ AND Ru(dipy),(CN), where M(nQ) means a micelle containing n quenchers.With this modification the kinetic equation for quenching is: In I / I , = - (k, + A[Q]) t - B[Q] [ 1 - exp ( - Ct)] where I, and I are the emission intensities at t = 0 and t, C = k , + ke[M] + k- and [Q] is the total concentration of quencher present. In the derivation of this equation it is assumed that in a micelle with n quenchers, the quenching rate constant is n times that with one quencher, as indicated above. In the general case eqn (1) gives a logarithmic decay curve with two distinguishable parts, viz. an initial faster decay determined by the second term, followed by a portion linear in time with the form of the first term. This case arises when the quenching constant k, is high w.r.t.the exchange of the quencher, so that the emitters in micelles already containing quenchers are rapidly quenched, the remainder having to await the equilibrium of quenchers among the micelles. Several examples of this behaviour have been r e p ~ r t e d . ~ If k, is sufficiently large the initial decay may be too fast to be observed in which case ‘static quenching’ apparently occurs, i.e. there is an apparent decrease in the initial emission intensity followed by the logarithmic decay term. The quenching of excited Ru(dipy)i+ by methyl anthracene was reported to show this static quench- ing phen~menon,~ but recent work5 with better time resolution has shown the apparent decrease in initial intensity to be a fast initial decay. A third variation occurs if k, is small and the second term of eqn (1) becomes negligible.In this case the decays are exponential and show Stern-Volmer behaviour, the Stern-Volmer constants varying with the micelle concentration. Many examples of this have been r e p ~ r t e d ~ ? ~ and the situation arises from the fact that the slow quenching allows equilibration of the quencher among the micelles. It has recently been reported7 that even at sub-micellar surfactant concentrations the quenching of charged emitters is considerably enhanced. This has been attributed to the formation of emitter-surfactant clusters which behave like micelles in that they have an affinity for the quencher. In the present work we have examined the quenching of excited Ru(dipy):+ and Ru(dipy),(CN), by Cu2+ and methyl viologen (MV2+) in the presence of sodium dodecyl sulphate (SDS) at concentrations above and below the critical micelle concentration (c.m. c .). EXPERIMENTAL Ru(dipy)i+ was precipitated as the perchlorate from a solution of the chloride (G. F. Smith). Ru(dipy),(CN), was kindly supplied by the Ciamician Institute, University of Bologna, and used without further treatment. Cu2+ was added as the sulphate (analytical reagent quality), MV2+ as the chloride (B.D.H.) and SDS was B.D.H. specially purified grade. Stock solutions of the latter were freshly made for the day of their use. Solutions of the mixed reagents were used immediately after making up in triply distilled water and were deaerated using pure nitrogen. Concentrations of Ru(dipy):+ between 20 and 150 pmol dmP3 and of Ru(dipy),(CN), of ca. 20 pmol dm-3 were used.Experiments were done at room temperature, i.e. 295 & 2 K.S. J. ATHERTON, J. H. BAXENDALE AND B. M. HOEY 2169 Emissions were generated using 530 nm 15 ns light pulses from a System 2000 J.K. neodymium laser, frequency-doubled and moni tored at 680 nmusing a standard monochromator- photomultiplier arrangment feeding a Tektronix 79 12A digitiser. The latter was controlled by a Commodore PET microcomputer which was also used to store the data on disk, to analyse the data and to give hard copies of the primary signals and functions derived from them. It has been showns that in some conditions high exciting light intensities can lead to anomalous kinetics for emission decays of Ru(dipy)g+ in SDS solutions.Care was taken to ensure that the intensities used were such that these effects were absent. RESULTS AND DISCUSSION QUENCHING OF EXCITED Ru(dipy)t+ BY Cu2+ ABOVE THE C.M.C. The decay constant for Ru(dipy)t+ in aqueous solution, 1.6 x lo6 s-l, is well established and we find the same value in SDS solutions below the c.m.c. However, at 12 mmol dm-3 SDS and higher we find the lifetime increases slightly and k = 1.25 x lo6 S-'. The Cu2+ quenching reaction in simple aqueous solution has been showng to be an electron transfer *Ru(dipy):+ + Cu2+ + Ru(dipy):+ + Cu+ with rate constant 7.7 x lo7 mol-1 dm3 s-l, and it was also shownlo that the quenching rate increases in the presence of 0.02 mol SDS. 8 - 1 I I I 0 2 I 6 8 [Cu2+] /mrnol drn-' FIG. 1 .-Quenching of Ru(dipy)i+ by Cu2+.Variation of quenching constant, kobsr with Cu2+ concentration. Numbers on lines are the SDS concentrations in mmol dm-3. Using SDS concentrations from 12 to 40 mmol dm-3 we have measured the effect of Cu2+ up to concentrations which increased the emission decay approximately fourfold at each SDS concentration. The decays were all first order over at least 90% of the reaction and the decay constants, kobs, were found to be linear with Cu2+ concentration, as shown in fig. 1. The general behaviour is consistent with the third variation of eqn (1) mentioned above when k , < k,[M]+k- and is small, in which case the first term is dominant and eqn (1) becomes In I l l , = - kobs t = - {k, + k,K[Cu2+]/( 1 + K[M])) t . (2)2170 QUENCHING OF FLUORESCENCE FROM Ru(dipy):+ AND Ru(dipy),(CN), The measured quenching constant, kobs, should thus be linear with [Cu2+] as in fig.1 giving lines of gradient S where 1 / S = (1 + flM])/k, K . Calculating [MI using c.m.c. = 8.2 mmol dm-3 and aggregation number 62 the plot in fig. 2 shows this to be true and from the line in fig. 2 we obtain k, = 2.1 x lo5 s-l and K = 2.0 x lo4 mol-1 dm3. The latter compares with 6 x lo4 mol-1 dm3 obtained from pyrene quenching3d and 1.0 x lo4 mol-1 dm3 from methyl pyrene quenching.3c 100 200 300 400 500 [ M I /pmol dm-3 FIG. 2.-Quenching of Ru(dipy)i+ by Cu2+ and MV". Variation of S, the gradients of lines in fig. 1 and 3, with micellar concentration. QUENCHING OF EXCITED Ru(dipy):+ BY MV2+ ABOVE THE C.M.C. This is also an electron-transfer reaction in simple aqueous solution *Ru(dipy):+ + MV2+ -+ Ru (dipy):+ + MV+ with6 rate constant 1.8 x lo8 mol-1 dm3 s-l.Measured at one MV2+ concentrat' n (2 mmol dmP3) and SDS concentrations up to 82 mmol dm-3, the rate constant was found in earlier work to be much increased at low concentrations of SDS and fell as the concentration of SDS increased.6 We have measured the decay in SDS concentrations up to 40 mmol dm-3 with MV2+ concentrations which give an increase of up to approximately fourfold in the decay rate. In these conditions the decays were found to be first order over at least 90% of the reaction and the first-order constants, kobs, were linear with [MV2+], as shown in Treating the data in the same way as for Cu2+, the plot of 1/S (from fig.3) against [MI shown in fig. 2 gives k, = 6.6 x lo5 s-l and K = 7 x lo4 mol-l dm3, although the latter could be in considerable error since it is determined by the small intercept in fig. 3. fig. 2.S. J. ATHERTON, J. H. BAXENDALE AND B. M. HOEY 2171 [ MV2'l/mmol dm-' FIG. 3.-Quenching of Ru(dipy)i+ by MV2+. Variation of quenching constant, kobs, with MV2+ concen- tration. Numbers on lines are SDS concentrations in mmol drnp3. QUENCHING OF EXCITED Ru(dipy),(CN), BY MV2+ ABOVE THE c.M.C. *Ru(dipy),(CN), + MV2+ -+ Ru(dipy),(CN)i + MV+ and the rate constantll is found to be 5.3 x lo9 mol-1 dm3 s-l. No experiments are reported with MV2+ in the presence of SDS. Several sources11$ l2 have given k = (3.8 f 0.1) x lo6 s-l for the excited-state decay constant in water and we have confirmed that this value also holds in SDS solutions up to 6 mmol dm-3, but at 16 mmol dm-3 and higher the decay constant decreases to (2.220.1) x lo6 s-l.At these latter SDS concentrations we have shown by pulse-radiolysis experiments, in which the reaction of e;, with the complex was followed, that > 96% of the complex is in SDS micelles. The longer lifetime in the micelles is paralleled by that in methano112 where k = 2.5 x lo6 s-l and presumably the less polar environment is responsible. The quenching kinetics in micellar solutions are not first order like those for Ru(dipy)g+, but over a range of SDS and MV2+ concentrations the logarithmic decay plot has fast and slow components (see fig. 4), as has been found for example in the quenching of pyrene by methylene iodide,3a ~ y r e n e ~ ~ by Cu2+, methyl p ~ r e n e ~ ~ by Cu2+ and Ru(dipy)g+ by methyl anthra~ene.~ Qualitatively this phenomenon is observed when the quencher is strongly bound to the micelles and has a high quenching rate constant so that there is rapid quenching in micelles containing both emitter and quencher followed by the slower emission from micelles containing emitter only.In simple aqueous solution this also is quenching by electron transfer In terms of eqn (l), when k, % k-+k,[M] this reduces to In I / I , = fi[exp ( - k, t) - 11 ( k , + B(k- + k,[M])) t (3 4 where fi = K[Q]/( 1 + K[M]), the average number of quenchers occupying a micelle. Analysing the slow components using only the latter part of the decays (fig.4) we find that these have decay constants within 10% of k, so that ~ ( k - + k,[M]) is negligible (3 b) and hence In I / I , = R[exp ( - k, t) - 11 - k , t an equation first derived in this form by Rodgers and Wheeler.362172 QUENCHING OF FLUORESCENCE FROM Ru(dipy)i+ AND Ru(dipy),(CN), 5*43 h 4.60 g 3 . 7 7 \ 5 E 4 2 . 9 5 2.12 110 308 506 70 4 tlns FIG. 4.-Logarithmic decay of emission from solution containing 20 pmol dm-3 Ru(dipy),(CN),, 500 pmol dm-3 MV2+ and 30 mmol dme3 SDS excited at 530 nm observed at 680 nm. 0 1 2 3 4 5 6 7 [MV2*l/10-' mmol dm-3 FIG. 5.-Quenching of Ru(dipy),(CN), by MV2+. Variation of extrapolated value of In Z/I, to t = 0 [ = N by eqn (3 c)] with MV2+ concentration. It will be seen from this equation that the form of the slow component is In I / I , = -Ha-k,t (3 4 so that extrapolation to t = 0 gives values of A{ = K[MV2+]/( 1 +KIM])}. We obtained values of A in this way for 40 mmol dm-3 SDS using 100-600 pmol dm-3 MV2+ and these are plotted against [MV2+] in fig.5. The inverse of the slope of this line, which should be given by (1 + K[M])/K, is 550 50 pmol dmF3. Using accepted values of 8 mmol dm-3 for the c.m.c. and 60 for the micelle aggregation number, [MI = 530pmol dm-3, the slope to be expected if K[M] B 1. The value 7 x lo4 mol-1 dm3 obtained above for K is entirely consistent with this, i.e. the MV2+ is entirely micellised in these conditions. We have also analysed the fast initial portion of the decays by computer fitting aS. J. ATHERTON, J. H. BAXENDALE A N D B. M. HOEY 2173 TABLE VA VALUES OF k , FOR QUENCHING OF Ru(dipy),(CN), BY MV2+ [SDS]/mmol dm+ [MV*+]/pmol dm-3 kq/107 S-1 20 20 20 30 30 30 40 40 40 200 300 400 200 400 500 200 400 600 1.09 & 0.12 1.21 kO.11 1.12 f 0.1 1 1.55k0.25 1.21 k0.09 1.07 +O.13 1.65 _+ 0.25 0.93 k0.22 1.18 k0.15 Values of k, obtained by computer fit of fast component (see fig. 4) to eqn (3b). value of k, to eqn (3b) using the experimental value of k,. The results are given in table 1, from which it can be seen that k, = 1.2 x lo7 s-l satisfies a range of MV2+ and SDS concentrations. QUENCHING OF EXCITED Ru(dipy),(CN), BY cu2+ ABOVE THE C.M.C. *Ru(dipy),(CN), + Cu2+ -+ Ru(dipy),(CN)z + Cu+ for which12 k = 3.9 x lo8 mol-I dm3 s-l. No work on quenching in the presence of SDS has been reported. For SDS concentrations from 20 to 40 mmol dm-3 and Cu2+ concentrations up to 150 pmol dm-3, we find that over at least 90% of the decay the reaction is first order and the rate constants increase slightly from 2.5 x lo6 s-l as the Cu2+ concentration increases but by only ca.20% at the highest Cu2+ concentration. However, at each SDS concentration the emission intensity at t = 0 decreases appreciably as the Cu2+ concentration is increased, as shown in fig. 6, i.e. there is 'quasi-static quenching'. This is clearly a system which follows eqn (3a) above, where k, is so large that the fast component cannot be observed with the time resolution of our equipment, and the n(k- + k,[M]) term is not entirely negligible w.r.t. k,. In these circumstances In I / I , = - A - { k, + n(k- + k,[M])} t In simple aqueous solution this is oxidative quenching and A gives the extent of quasi-static quenching.The data are not sufficiently accurate to allow a meaningful analysis to obtain k- and k, as has been done for other but n = K[Cu2+]/(1 +K[M]) can be obtained from the intercepts of the In I / I , against t plots. Examples of these are given in table 2 for 20 mmol dm-3 SDS. Compared with MV2+ in the same conditions Cu2+ is much more efficient, requiring only ca. 20% of the MV2+ concentration to produce the same extent of the fast quenching. We see, however, that the MV2+ observations are understandable if the MV2+ is almost entirely micellised and, in terms of the model used, no matter how big the value of k, this should give the maximum extent of quasi-static quenching possible with any quencher. Moreover, the maximum value of A( = K[Q]/(l +K[M])} which can be attained is [Q]/[M], whereas it will be seen from table 2 that for each Cu2+ concentration the experimental value is approximately three times this.This pattern is followed for 25, 30 and 40 mmol dm-3 SDS solutions. It is clear that far more Cu2+ is associated with the emitter on the2174 QUENCHING OF FLUORESCENCE FROM Ru(dipy):+ AND Ru(dipy),(CN), 0 0.2 0.4 0.6 0.8 tlns FIG. 6.-Quenching of Ru(dipy),(CN), by Cu2+. Logarithmic decay of 20 pmol dm+ emitter in 25 mmol dm-3 SDS showing quasi-static quenching. The numbers on the lines are Cu2+ concentration in pmol dmP3. TABLE 2.-QUASI-STATIC QUENCHING OF Ru(dipy),(CN), BY CU2' ~~~~~~ ~ ~~~ [Cu2+]/pmol dm-3 15 20 30 50 60 ECu2+l/[M1 0.079 0.105 0.158 0.263 0.315 A 0.40 0.41 0.61, 0.71 0.88 1.09 Solution contains 20 mmol dm-3 SDS. A = In I / & at t = 0 obtained from data as in fig.6 . [MI calculated using c.m.c. = 8.2 mmol dm-3 and aggregation number 62. micelle than can be accounted for by the simple model and some more specific affinity for the emitter must be evoked. There is evidence for this in simple aqueous solution12 based on abnormally efficient kinetic quenching and an association constant of 200 mol dm-3 has been derived. There can be no similar association in solution in the present system since little Cu2+ or emitter are in solution. However, it is not unreasonable to suppose that this affinity may cause micelles containing the emitter to have a higher affinity for Cu2+ than the unoccupied micelles.We then would have two equilibria: Cu2++M$CuM K Cu2+ + RUM + CuRuM KR where CUM and RUM represent micelles containing Cu2+ and emitter, respectively.S. J. ATHERTON, J . H. BAXENDALE AND B. M. HOEY 2175 From these equilibria we have: By making some approximations we will now use this equation to relate the extent of the emission left after static quenching, I,, to the quencher concentration. We will neglect multiple occupancy of the micelles which is reasonable in view of the low concentrations of Cu2+ and emitter used in these experiments. Also, we will assume that there is no Cu2+ free in solution which, from the micelle association constant, will be approximately the case. Now from the extent of static quenching we can calculate how much of the emitter is associated with Cu2+.Thus from table 2, 60 pmol dmP3 Cu2+ quenches 66% of the emitter, i.e. Cu2+ is associated with 13 of the 20 pmol dm-3 emitter present. Assuming single occupancy this means that ca. 13 % of the Cu2+ is present in the species CuRuM. A similar extent is found over the range of Cu2+ concentrations. In view of the complexity which is necessary to take this into account, we shall assume that this is negligible and equate [CUM] to the total Cu2+ present. With these approximations and remembering that Ir/Io = [RuM]/([RuM] + [RuCUM]) the above equation leads to Io/Ir = KR[Cu2+]/K[M] + 1 (4) where [Cu2+] is the added Cu2+. Fig. 7 shows that I J I , is reasonably linear with [Cu2+] for a range of SDS concentrations, and fig.8 shows that the reciprocal slopes of these lines are linear with the calculated micellar concentration. From fig. 8 we obtain K,/K = 6.3, which is a measure of the preference of Cu2+ for micelles containing the emitter. QUENCHING OF Ru(dipy):+ BY MV2+ BELOW THE C.M.C. MV2+ is an even more efficient quencher at SDS concentrations below the c.m.c. We have investigated the system using a 4 mmol SDS solution. At all concen- trations of Ru(dipy)$+ used, viz. 29- 114 pmol dm-3, the emission kinetics are first order over at least 90% of the decay and there is no quasi-static quenching. The MV2+ is more effective as a quencher the smaller the Ru(dipy)i+ concentration (fig. 9), and over the above range 20pmol dm-3 MV2+ causes substantial increases in the quenching rate constants.This compares, for example, with several hundred micro- molar required at 12 mmol dmP3 SDS. As for the micellar solutions, rate constants are linear with MV2+ concentration, as shown in fig. 9. There is clearly an association of MV2+ and the emitter, as is the case for micellar solutions, but at 4 mmol dmP3 SDS there can be no micelles of the conventional type. However, it has been shown7 that at 4 mmol dmP3 SDS there is an association of Ru(dipy)t+ with SDS ions to form a cluster containing several Ru(dipy)$+, and that very efficient quenching occurs with methyl anthracene, in this case quasi-static. Those results were explained quantitatively by treating the cluster as a micelle able to associate with methyl anthracene in the manner outlined above.It seems likely that MV2+ behaves similarly. However, here there is no quasi-static quenching and the kinetics are of the same type described above for MV2+ in micellar solutions, which originate from a rapid equilibrium of quencher with micelles and a relatively small quenching constant k,.2176 QUENCHING OF FLUORESCENCE FROM Ru(dipy)$+ AND Ru(dipy),(CN), FIG. 7.-Quenching of Ru(dipy),(CN), by Cu2+. Variation of residual emission after quasi-static quenching, I,, obtained from data as in fig. 4 with Cu2+ concentration. The numbers on the lines are SDS concentrations in mmol dm-3. [Ml/10-' mmol dm-3 FIG. &-Quenching of Ru(dipy),(CN), by Cu2+. Variation of gradients, S, from fig. 7 with micellar concentration to compare with eqn (4).S. J.ATHERTON, J. H. BAXENDALE AND B. M. HOEY 2177 9- I - 1 - I I 1 I 0 5 10 15 20 [ MV2+] /pmol dm-3 MV2+ concentration. Numbers on lines are Ru(dipy)z+ concentrations in pmol dm-3. FIG. 9.-Quenching of Ru(dipy)i+ by MV2+ below the c.m.c. Variation of quenching constant, kobs, with 1 1 I 1 1 I I 0 LO 80 120 [Rul /pmol dm-’ FIG. 10.-Quenching of Ru(dipy)g+ by MV2+ below the c.m.c. Variation of gradients of lines in fig. 9 with Ru(dipy)i+ concentration at 4 mmol dm-3 SDS. Formulating this in terms of a Ru(dipy)i+-SDS cluster C, we have C + MV2+ e (CMV2+) K,. Assuming that multiple occupancy by MV2+ can occur this gives which is of the same form as eqn (2) above. Hence the observed quenching constants should be linear with [MV2+] as found (fig.9) and the slopes of the lines are given by k , Kc/( 1 + K,[C]). Since the concentration of SDS is large compared with that of Ru(dipy)E+, it would be expected that the composition of the clusters would not change2178 QUENCHING OF FLUORESCENCE FROM Ru(dipy),2+ AND Ru(dipy),(CN), and that the cluster concentration would be proportional to the concentration of Ru(dipy)i+. Hence, we should expect that the reciprocal of the gradients of the lines in fig. 9 would be linear with Ru(dipy)g+ concentration. Fig. 10 shows this to be the case and from the line we may obtain k, and K,. We find K , = 1 . 1 x lo5 mol-l dm3 and, if each Ru(dipy)g+ gives rise to only one cluster, k, = 2.6 x lo7 s-l. However, previous experiments have indicated that as many as eight Ru(dipy);+ may be contained in one c l ~ s t e r , ~ in which case we have k, = 3.3 x lo6 s-l and K, = 8.8 x lo5 mol-1 dm3. These values compare with 6.6 x lo5 s-l and 7 x lo4 mol-l dm3 obtained above for the micellar system.The higher value of k, is perhaps to be expected since it is likely that the clusters are smaller than micelles and the emitter and quencher are in closer proximity as a result. The higher value of K, suggests that the structure of the clusters differs from that of micelles. 0 0.5 1 .o 1.5 [Cu2+]/mmol dm-3 FIG. 1 1 .-Quenching of Ru(dipy)i+ by Cu2+ below the c.m.c. Variation of quenching constant, kobs, with Cu2+ concentration at 4 mmol dmP3 SDS. e, x , 0 and + are values at 29, 57, 86 and 114 pmol dmP3 Ru(dipy)i+, respectively. QUENCHING OF Ru(dipy)i+ BY c U 2 ' BELOW THE C.M.C.Cu2+ also quenches in these conditions but it is not as effective as MV2+. Thus at 6 mmol dm-3 SDS concentrations, ca. 1 mmol dm-3 Cu2+ are required to give appreciable quenching (20 pmol dm-3 for MV2+). Nevertheless for Cu2+ this is approximately five times more effective than at 20 mmol dm-3 SDS. The kinetics are also different from those of MV2+ quenching, for although the emission decays are first order over at least 90% of the reaction and there is no quasi-static quenching, the first-order constants are not linear with Cu2+ concentration, as shown in fig. 11. Instead the quenching constants appear to approach a limiting value and moreover are seen to be independent of the Ru(dipy)i+ concentration over the range 29-1 14 pmol dm-3.Again there must be association of quencher and emitter since the rates are well in excess of those for simple aqueous solutions, but the concentration dependence suggests that it is not the micellar or cluster type of association encountered so far in this work. The approach to a limiting value of the quenching constant suggests thatS. J. ATHERTON, J. H. BAXENDALE AND B. M. HOEY 2179 it is a 1 : 1 association of emitter and quencher rather than the micellar type which allows the emitter to be associated with many quenchers. The reason for this difference is probably due to the fact that, unlike MV2+, Cu2+ is strongly associated with SDS below the c.m.c. The evidence for this, presented in detail below, is that Cu2+ is a very efficient quencher for excited Ru(dipy),(CN), below the c.m.c.whereas MV2+ is no more effective than in simple aqueous solution. We conclude that Cu2+, like Ru(dipy):+, forms a micellar type species, which can take up Thus with Cu2+ and Ru(dipy)!+ neither emitter nor quencher exist as simple ions in SDS solution and the quenching observations are consistent with the association of the two SDS-complexed ions C and R to form CR: Ru(diPY),(CN)z. C + R e C R KR. 0 5 10 [Cu2'] /mmol dm-3 15 FIG. 12.-Quenching of Ru(dipy)g+ by Cuz+ below the c.m.c. Variation of quenching constant, kobs, with Cu2+ concentration plotted to compare with eqn ( 5 ) . SDS concentration 6 mmol dm-3, 0, (data obtained from smoothed observations given in fig. 1 1 at various emitter concentrations), and 4 mmol dm-3, x , (using 114 pmol dm-3 emitter).The quenching occurs in the species CR, and since there is no static quenching, this equilibrium must be rapidly established within the lifetime of the excited emitter. In this situation it is readily shown that if the lifetime of the emitter in the species CR is Ilk, and the Cu2+ concentration is such that C % R then where ko is the natural lifetime of the emitter. tration but is given by the equation The observed quenching constant kobs is no longer linear with quencher concen- kobs - k O = (kq - k o ) KR[cl/(l + KR[c]). ( 5 ) If [C] is proportional to [Cu2+] then 1 /(kobs - k,) should be linear with 1 /[Cuz+]. As shown in fig. 12 this is the case for 6 mmol dm-3 SDS and a range of Ru(dipy):+2180 QUENCHING OF FLUORESCENCE FROM Ru(dipy):+ AND Ru(dipy),(CN), concentrations and for 4 mmol dm-3 SDS with 114 mmol dm-3 Ru(dipy):+.If C contains only one Cu2+ the lines in fig. 12 give k , = 8.7 and 10.3 x lo6 s-l and KR = 4.5 and 4.1 x lo3 mol-1 dm3 for 6 and 4 mmol dm-3, respectively. If, as seems likely, C contains more Cu2+, then KR is proportionately larger. Note that k, is approximately fifty times the value obtained with the micellar systems, which again may be due to the closer proximity of quencher and emitter and/or the presence of more than one Cu2+ in C. QUENCHING OF Ru(dipy),(CN), BELOW THE C.M.C. Unlike Ru(dipy)i+, below the c.m.c. Ru(dipy),(CN), does not show an increased sensitivity to quenching by MV2+, which is consistent with our conclusion that Ru(dipy):+ as a consequence of its charge forms micelle-like aggregates by association 0 5 10 15 20 [ Cuz+l /pmol dm-3 FIG.13.-Quenching of Ru(dipy),(CN), by Cu2+ below the c.m.c. Variation of I, the initial intensity remaining after quasi-static quenching obtained from data as in fig. 6, with CU*+ concentration. Plotted according to eqn (6). 0 and x are for 6 and 4 mmol dm-3 SDS, respectively. with SDS ions. The uncharged cyanide complex would not be expected to behave in this way. However, at 4 and 6 mmol dm-3 SDS, Cu2+ is still a very effective quencher in the concentration range below 20 pmol dm-3 and behaves kinetically as at SDS concentrations above the c.m.c. (see fig. 6), i.e. there is quasi-static quenching followed by exponential decay with the unquenched aqueous rate constant 3.7 x lo6 s-l.At 20 pmol dmP3 Cu2+ the quasi-static quenching removes > 70% of the initial emission and over a range of Cu2+ concentrations there is little difference between 4 and 6 mmol dm-3 SDS. This again indicates a strong affinity of the emitter for Cu2+ but clearly at these low concentrations it cannot involve the simple ion Cu2+ referred to above.12 We suggest that it is the association of the emitter with a Cu2+-SDS cluster, the latter behaving as though it were a micelle. We then have the equilibrium C + E e R E K , where C is the cluster and E the emitter. In the associated species RE there will be quasi-static quenching as in a micelle and all the emission originates in the free emitter,S. J. ATHERTON, J. H. BAXENDALE AND B. M. HOEY 2181 E.Assuming a micellar-type association with multiple occupancy of C by E then [El I 1 [E]+[RE] =lo= l+K,[C] where I is the initial emission intensity at a concentration [C] of the Cu2+-SDS species and I, is the intensity in the absence of Cu2+. If all the Cuz+ is present as C then I,/I should be linear with [Cuz+], and fig. 13 shows this to be reasonably so. The fact that there is little difference between 4 and 6 mmol dm-3 SDS supports the assumption that the Cu2+ is present entirely as C. From fig. 13, we find KE = 1.25 x lo5 mol-1 dm3. If C contains more than one Cu2+ then K , is proportionately larger. Note that the absence of similar behaviour by MV2+ indicates that the latter is unable to form micelle-like clusters in these conditions. Note also that similar association of SDS with multiply charged metal ions in solution is to be expected and that as a result the ion-micelle association constants usually obtained do not necessarily refer to the free ion in solution.It would seem that the high charge density of the small metal ions is required for such association, since MV2+ does not behave like Cu2+ in the above conditions. CONCLUSIONS To summarise, we find the observations are explicable by Infelta's general kinetic equation' and in particular : (a) The enhanced quenching of the emission from Ru(dipy)i+ by Cu2+ and MV2+ in micellar SDS solutions is accounted for quantitatively in terms of a completely micellised emitter and rapid equilibration of the distribution of the quencher between aqueous and micellar phases.The quenching is slow compared with the equilibration. (6) In the quenching of Ru(dipy),(CN), emission by Cu2+ and MV2+ in the same conditions, the quenching rates are faster than the equilibration. Cu2+ has a specific affinity for micelles containing the emitter. (c) The quenching of Ru(dipy)i+ emission by Cu2+ and MV2+ is still enhanced by concentrations of SDS below the c.m.c. ( d ) The quenching of Ru(dipy),(CN), emission by MV2+, in contrast to that by Cu2+, is not enhanced by SDS below the c.m.c. The emitter seems to associate with Cu2+-SDS clusters as it does with micelles. Values of the quenching and association constants, k, and K are given in table 3. TABLE 3.-vALUES OF QUENCHING AND ASSOCIATION CONSTANTS cu2+ MV2+ k,/s-' Klmo1-l dm3 kq/ss1 Klmo1-l dm3 micellar systems Ru(dipy):+ 2.1 x 105 2.0 x 104 6 . 6 ~ lo5 7~ 104 Ru(dipy),(CN), large 6.3a 1.2x 107 large Ru(dipy)z+ 9.5 x 106 4.3 x 104b 3.3 x 106 8.8 x Ru(dipy),(CN)z large 1.25 x below the c.m.c. - - a Ratio of association constants for Cu2+ and micelles with and without emitter. K for K for association of Ru(dipy), (CN), association of quencher with Ru(dipy)t+-SDS cluster. with Cu2+-SDS cluster.2182 QUENCHING OF FLUORESCENCE FROM Ru(dipy)t+ A N D Ru(dipy),(CN), We thank Dr E. J. Land for the use of the laser at the Paterson Laboratory, Christie Hospital, the Royal Society for financial support for the development of our computerised data acquisition and analysis facility, and Mr P. R. Baxendale who interfaced the Tektronix 7912A digitiser to the Commodore PET and con- tributed substantially to the programming. P. P. Infelta, Chem. Phys. Lett., 1979, 61, 88. J. C. Dederen, M. Van der Auweraer and F. C. De Schryver, Chem. Phys. Lett., 1979, 68, 451. (a) P. P. Infelta, M. Gratzel and J. K. Thomas, J. Phys. Chem., 1974, 78, 190. (b) M. A. J. Rodgers and M. F. S. Wheeler, Chem. Phys. Lett., 1978,53, 165. (c) J. C. Dederen, M. Van der Auweraer and F. C. De Schryver, J. Phys. Chem., 1981,85, 1198. ( d ) F. Grieser and R. Tausch-Treml, J. Am. Chem. SOC., 1980, 102, 7258. N. J. Turro and A. Yekta, J. Am. Chem. SOC., 1978, 100, 5951. M. A. J. Rodgers and J. H. Baxendale, Chem. Phys. Lett., 1981,81, 347. M. A. J. Rodgers and J. C. Becker, J. Phys. Chem., 1980, 84, 2762. ’ J. H. Baxendale and M. A. J. Rodgers, Chem. Phys. Lett., 1980, 72, 424. U. Lachish, M. Ottolenghi and J. Rabani, J. Am. Chem. Soc., 1977, 99, 8062. D. Meisel, M. S. Matheson, W. A. Mulac and J. Rabani, J. Phys. Chem., 1977, 81, 1449. lo D. Meisel, M. S. Matheson and J. Rabani, J. Am. Chem. SOC., 1978, 100, 117. l 1 G. L. Gaines, J. Am. Chem. Soc., 1979, 83, 3088. l2 J. N. Demas, J. W. Addington, S. H. Peterson and E. W. Harris, J. Phys. Chem., 1977, 81, 1039. (PAPER 1 / 1403)