4400
J. Phys. Chem. 1982, 86,4400-4404
Photophysics of Indole in Tetramethyisilane Solution. I s There a Photoionization Threshold? R. Kleln,' J. P. Balllnl, and I. Tatlscheff Instltut Curie, LA 198,associi au CNRS et 2 l'Universit6 P. et M. Curie, Section de Physique et Chimie, 7523I Paris C a e x 05,France (Received: Februafy 9, 1982;In Final Form: June 29, 1982)
Indole in tetramethylsilane(Me&) solution shows a 64% drop of the fluorescence quantum yield with decreasing excitation wavelength. In degassed solution r#Jf = 0.50 f 0.05 at wavelengths higher than 240 nm and r#Jf = 0.18 f 0.02 at wavelengths lower than 225 nm. The fluorescence lifetime, independent of excitation wavelength, is 9.4 f 0.1 ns. Dynamic quenching by NzO is observed, the fluorescence lifetime and the quantum yield being both 19% lower in the presence of NzO at atmospheric pressure than in degassed solution. The formation of N2 in the presence of NzO is higher in the first absorption band (lLa,lLb) as measured at 292 and 254 nm, 4(N2) = 0.12 f 0.02, than in the second absorption band (lBb),r#J(N2)= 0.07 f 0.02 at 214 nm. From these results it is concluded that NzO cannot be confidently used to determine the photoionization threshold of indole in Me4%, contrary to previous claims.
bond dissociationof indole in hydrocarbon s ~ l v e n t s . ~Up J~ Introduction to now, neither fluorescence data nor quenching data were Among different nonpolar solvents, tetramethylsilane known for indole in Me4Si. The present work deals with (Me4Si)was shown to be the most efficient one to decrease the fluorescence of indole in Me4Si and the effect of N20. the photoionization threshold of tetramethylparaResults are compared with similar ones for indole in cyphenylenediamine (TMPD).'v2 Me4Si was then selected clohexane solution. The determination of a photoionizain order to study the photoionization threshold of indole tion threshold for indole in Mel% using NzO as electron in condensed phase with two methods: photoconductivity scavenger is discussed in the light of these results. measurements and formation of N:, from electron scavenging by Nz0.3 The photoionization threshold for inMaterials and Methods dole in Me4& solution was found by both methods at 4.95 The solutions were made with Me4% (Merck), vacuum eV (250 nm).3 In spite of the good agreement of the two distilled over molecular sieves, or with cyclohexane methods, this result remains ambiguous. Firstly, the de(Merck), percolated through a column of activated Woelm termination of a photoionization threshold by photoconsilica. Indole (Calbiochem) was used as received. NzO (Air ductivity largely depends on the sensitivity of the meaLiquide) was purified by distillation. Degassing was obsurement device as recently e ~ i d e n c e d . ~Secondly, the tained by five successive freeze-thaw cycles. determination of photoionization yield using the N20 Corrected emission spectra, excitation spectra, as well method of Jortner5 relies, in the present case, on two deas fluorescence quantum yields were obtained as earlier batable assumptions: that the nitrogen formation is only describedloJ3except for the following improvements: use due to the electron scavenging and that NzO does not of an Ortec 9315 photocounter, data treatment with a quench indole fluore~cence.~ According to Lumry,6 N 2 0 Hewlett-Packard 9835 microcomputer, and addition of a can attack excited states of indole. In fact, concerning an third EM1 6256 photomultiplier aligned with the excitation eventual quenching effect with Nz formation we can beam for absorption measurement. A film of sodium mention that for indole in cyclohexane solution N2 was salicylate pulverized on a glass window was used as a found by exciting with wavelengths higher than 265 nm;' quantum counter before the monitoring and absorption this was shown to result from a quenching reaction bephotomultipliers. The absorption spectra were also meatween N20 and indole in its first singlet excited ~ t a t e . ~ , ~sured with a Cary 118CX spectrophotometer. The Moreover, indole shows a drop of the fluorescence quanfluorescence quantum yield was determined from a cortum yield in the second absorption band (lBb), which is rected emission spectrum of indole in degassed Me4& solvent dependent.'O This effect was shown to be due to compared to the one of p-terphenyl in degassed cyclothe photoionization in aqueous solutionsgJ1and to the N-H hexane used as a standard with = 0.9314 taking into account the correction, (n/n,,J2,for the refractive index of both solvents using for Me4% nref= 1.358 and for cy(1) R. A. Holroyd and R. L. Russel, J. Phys. Chem., 78, 2128 (1974). clohexane n = 1.426. Fluorescence lifetimes were measured (2) K. C. Wu and S. Lipsky, J. Chem. Phys., 66, 5614 (1977). by single-photon counting with synchrotron radiation, (3) A. Bernas, D. Grand, and E. Amouyal, J . Phys. Chem., 84, 1259 (1980);D. Grand, A. Bernas, and E. Amouyal, Chem. Phys., 44,73 (1979). LURE (Orsay),using the apparatus and method described (4) K. Siomos, G. Kourouklis, and L. G. Christophorou, Chem. Phys. in ref 15. Photolysis experiments were performed with Lett., 80, 504 (1981).
(5) J. Jortner, M. Ottolenghi, and G. Stein, J . Phys. Chem., 66, 2037 (1962). ( 6 ) R. Lumry, Photochem. Photobiol., 27, 819 (1978). (7) T. R. Hopkins and R. Lumry, Photochem. Photobiol., 15, 555 (1972). (8) R. Klein, I. Tatischeff, M. Bazin, and R. Santus, J. Phys. Chem.. 85, 670 (1981). (9) J. Zechner, G. Kohler, N. Getoff, I. Tatischeff, and R. Klein, Photochem. Photobiol., 34, 163 (1981). (IO) I. Tatischeff and R. Klein, Photochem. Photobiol., 22, 221 (1975).
(11) H. B. Steen, M. K. Bowman, and L. Kevan, J. Phys. Chem., 80, 482 (1976). (12) I. Tatischeff, R. Klein, T. Zemb, and M. Duquesne, Chem. Phys. Lett., 54, 394 (1978). (13) I. Tatischeff and R. Klein in "Excited States of Biological Molecules', J. B. Birks, Ed., Wiley, New York, 1976, p 375. (14)I. B. Berlman in "Handbook of Fluorescence Spectra of Aromatic Molecules", Academic Press, New York, 1971, p 220. (15) J. P. Ballini, M Daniels, and P. Vigny, J. Lumin., in press.
0022-3654/82/2086-4400$01.25/0 8 1982 American Chemical Society
The Journal of Physical Chemistty, Vol. 86, No. 22, 1982 4401
Photophysics of Indole in Tetramethyisilane
WAVELENGTHCnm) :
.
35 0, :
.
.
.
30 8, :
:
:
:
:
250 ~
:
\
\
w
l
LL
1
220
240
260
280
WAVELENGTH(^^) W A V E N U M B E R (prn-1)1
Figure 1. Absorption and fluorescence emission spectra of 3 X lo-' M Indole in degassed M,Si solution. Excitation wavelength: 260 nm. Monochromator bandwidth: emission, 1.6 nm; excitation, 5.2 nm.
three different light sources: a low-pressure mercury lamp (Pen-Ray) with a MTO interference filter (254 f 20 nm), a zinc lamp (931063 Philips) with a Matra interference filter (214 f 30 nm), and a Chromatix dye laser, used with rhodamine 6G, giving 25 pulses s-' with a fwhm of 1 p s at 292 nm. The laser intensity was monitored by a pyroelectric probe (RjP-735 from Laser Precision Corp.) with an energy meter (Rj-7100). Actinometry was performed with 6 X M potassium ferrioxalate assuming the same yield at the three wavelengths, &+,a+ = 1.25.16J7Absorbed dose rates were respectively (3.6 f 0.3) X 1015photons min-' at 254 nm, (1.3 f 0.1) X 10l6photons min-l at 214 nm, and (5 f 1) X 1015photons min-' at 292 nm. The saturation with N20 of Me4& solutions was obtained by two methods. In the first method, degassed Me4&, maintained at low temperature (0-5 "C), was shaken with N20 added step by step until reaching the atmospheric pressure. In the second method, an empty flask, evacuated at bar, was filled with N20 at controlled pressure, and then N20was transferred to the flask containing the degassed solution by freezing with liquid nitrogen. This later flask was in two parts: a Pyrex balloon and a Hellma parallelipedic quartz cell (10 X 10 mm). It was checked with blanks without indole that the Me4& solution saturated with N20by the fmt method did absorb at wavelengths lower than 230 nm (Figure 3). In the same wavelength range, the indole solution depicts an increasing absorption. As shown with the second method, the absorbance at 214 nm of N20 in Me4& was proportional to the measured pressure. The solubility of N 2 0 in Me4Si not being known, the increase of the absorbance at 214 nm was taken as a relative measurement of the N20 concentration. For photolysis experiments, 20 mL of M indole in degassed solution of Me4Si was saturated as above by N20 at atmospheric pressure and then deaerated by five successive freeze-thaw cycles with liquid nitrogen as the freezing agent. Samples were irradiated with total absorbed doses 11.2 X 1019 photons. The gases produced by photolysis were analyzed by gas chromatography (Fractovap Carlo Erba) with helium as a vector gas.
Results Fluorescence Data for Indole in Degassed Me4Si. Indole in degassed Me4& emits a fluorescence with a max(16)C. G. Hatchard and C. A. Parker, R o c . R. SOC. London, Ser. A, 235, 518 (1956).
(17)W. D.Wilson and J. F. Foster, Biochem. Biophys. Res. Commun., 38,552 (1970).
1
Figure 2. Absorption spectrum, fluorescence excitation spectrum, and fluorescence quantum yields of 3 X lo-' M indole in degassed Me,Si solution as a function of the excitation wavelength. (-) Absorption spectrum ("in situ'' measurement), S(X,) = 1 - 10-A(bywhere A(&) is the 1-cm absorbance at X, and I = 0.2 cm Is the light pathway in the fluorescence cell. (- -) Excitation spectrum normalized at 270 nm with respect to the absorption curve. Emission wavelength: 310 nm. Monochromator bandwidth: excitation, 1.2 nm; emission, 12.8 nm. (. .) Fluorescence quantum yields (4 ,).
-
TABLE I: Fluorescence Data for Indole in Tetramethylsilane and Cyclohexane Solutions in the Presence of N,O at Atmospheric Pressure or in Degassed Solution"
nm) nm) nm) nm) T ~ / Me,Si, degassed 0.50 0.18 0.36 9.5 9.3 0.98 Me,Si t N,O 0.40 0.145 0.36 7.7 7.6 0.99 cyclohexane, 0.49 0.15 0.31 8.6 8.7 1.01 degassed cyclohexane + 0.37 0.115 0.31 6.8 6.6 0.97 N,O a @ f , l , $I!,~: fluorescence quantum yields. T ~ 7 ,* : fluo. rescence lifetimes (x109s). Uncertainty: @ f , * 10%;T , r 0 . 1 ns. P = @ f , , / @ f , l .
T ~
imum at 3.35 pm-' (299 nm) (Figure l),close to that of indole in cyclohexane, 3.33 pm-' (300 nm). The emission spectrum is not the mirror image of the absorption spectrum (Figure 1). It reflects exclusively the 'Lb transitions, as expected in a nonpolar solvent,'* whereas the 'La and 'Lb electronic transitions overlap extensively in the absorption spe~trum.'~The first absorption band and the first emission band peak at the same wavenumber, 3.48 pm-' (Figure l),which corresponds to the 'Lb 0-0level in Me4Si. The difference between this peak and the fluorescence maximum v, - vf = 0.13 pm-', corresponds to the difference, 1310 cm-l, between the 'Lb 0-0 level and a 'Lb level found in the absorption spectra of indole by the solvent perturbation t e c h n i q ~ e . ' ~The + ~ ~difference v, vf taken as a measure of the Stokes shiftlo is the lowest value up to now found for indole in fluid solution.1° The fluorescence quantum yield obtained for indole in degassed Me4Si,with 265-nm excitation wavelength, was I#Jf = 0.50 f 0.05, very close to the one of indole in cyclohexane, 0.49.1° In Figure 2 are compared the absorption spectrum, the corrected excitation spectrum, and the fluorescence quantum yields of indole in degassed Me4Sias a function of the excitation wavelength. The excitation spectrum follows the absorption curve for wavelengths higher than 240 nm. The fluorescence quantum yield is constant in (18)H.Lami, J . Chem. Phys., 67, 3274 (1977). (19)E.H.Strickland,J. Horwitz, and C. Billupa, Biochemistry, 9,4914 (1970). (20)M. Martinaud and A. Kadiri, Chem. Phys., 28, 473 (1978).
4402
The Journal of Physical Chemistry, Vol. 86, No. 22, 1982
Klein et al.
-
Figure 3. Excitation spectra of indole 3 X M in Me4Si in the , and in degassed solution, normalized at 270 nm: (-) presence of NO in degassed solution; (. .) with N20at atmospheric pressure (A(W)zi4 ",,, = 0.2); (-'-) with NO, at pressure higher than atmospheric pressure (A(OD),i4 ",,, = 0.53); (- - -) absorption spectrum of Me,Si solution with N,O at atmospheric pressure.
.
0.80
0.81 0.81 0.82
0.81
i
0.01 cyclohexane
0.76
0.77
0.78
0.76
0.78
0.77 i
0.01
ot,,:
fluorescence q u a n t u m yield, fluorescence lifetime, and triplet q u a n t u m yield, respectively, in degassed solutions. of, T f , Ot : items in t h e presence of N,O a t atmospheric pressure. At 2 6 5 n m . Reference a Qf,o,
Tf,,,,
I/ L/
I
I
04 A(oD)i214n m i Figure 4. Stern-Volmer plot of F,IF - 1 vs. A(OD),,, nm. Excitation wavelength: (0)265 and (X) 215 nm.
TABLE 11: Quenching Effect of N,O for Indole in Me,Si and Cyclohexane Solutionsa
Me,Si
0.1
8.
the range 280-240 nm, 4f,l = 0.50 f 0.05, it drops between 240 and 225 nm, and it remains again constant in the range 225-210 nm, 4t,2= 0.18 f 0.02. The ratio 4f,2/4f:1= 0.36 is somewhat higher than that found for indole in cyclohexane, 0.31.1° The fluorescence lifetimes were determined by exciting either at 265 nm (71) or a t 215 nm (72) with the emission wavelength at 300 nm. The decay curves followed a single-exponential decay law. The fluorescence lifetimes of indole in degassed Me4Si and cyclohexane are reported in Table I, together with the fluorescence quantum yields. The ratios ~ ~ being / 7close ~ to 1 imply a fluorescence lifetime independent of the excitation Wavelength. The mean values were 9.4 f 0.1 ns in degassed MelSi and 8.7 f 0.1 ns in degassed cyclohexane, in rather good agreement with the value of 9.0 ns previously found in cyclohexane.21 Effect of N 2 0 . The shape of the fluorescence emission spectrum was identical in the presence of N 2 0 and in degassed solution. The only effect was a decrease of the fluorescence intensity in the presence of N20. The excitation spectrum in the presence of N 2 0 and that in degassed solution normalized at 270 nm were superposed from 290 to 220 nm (Figure 3). At wavelengths lower than 220 nm, the variation originated from light absorption by N20 as evidenced by calculating the fractional absorption + A2),where A, and of indole, B = Al(l - 10-(A1+A~)I)/(A1 Az are respectivelythe 1-cm absorbances of indole and N20 in Me4Siand I = 0.2 cm. The fluorescence quantum yields were lower in the presence of N 2 0 (4f) than in degassed solution ($qO)and the fluorescence lifetimes were shorter ) in degassed solution (7f,o) in the presence of N20 ( T ~ than (21) W. B. De Lauder and P. Wahl, Biochim. Biophys. Acta, 243, 153 (1971).
0.1
whatever the excitation wavelength both in Me,Si and in cyclohexane (Table I). The ratio 7f/7f,o parallels the ratio 4f/4f,o(Table 11). Taking the increase in 1-cm absorbance at 214 nm of the indole solution in Me4Si, A(OD), as proportional to the concentration of N 2 0 and to the molecular extinction coefficient of N 2 0 in Me4Si (e), the Stern-Volmer relation may be written Fo/F = 1 Ks,A(OD)/t
+
where Fo and F are respectively equal to the ratios Io/Bo and I / B , i.e., to the fluorescence intensities in the absence (Io)and in the presence (I) of N 2 0 divided by the absorption or fractional absorption in the absence (Bo)and in the presence ( B )of N20. The plot of Fo/F vs. A(0D) (Figure 4) is linear by exciting either at 265 nm, the wavelength for which B = Bo (no light absorption by N20), or at 215 nm, the wavelength for which B # B,. From the slope, identical at the two wavelengths, we can calculate the value Ksv/t = 1.25. The similarity of the SternVolmer curves obtained at either 265 or 215 nm (Figure 4) and the similar decrease at these two wavelengths of the fluorescence quantum yield and of the fluorescence lifetime in the presence of N 2 0 (Table 11) imply that only the relaxed fluorescent state is quenched by N20 in the present concentration range. The quenching quantum yield can be expressed by 4q = k,(N,O)/[kf + k,, + kq(N2O)I where k,, kf, and k , are respectively the rate constants for the quenching reaction and for the radiative and radiationless deactivation. Hence, the quenching quantum yield in the first absorption band, at 265 nm, in the presence of N20 at atmospheric pressure can be calculated from = 1 - 4f/4f,o = 1 - 7 f / 7 f , o taking the mean values of Table I1 as 0.19 f 0.02 in Me4Si and 0.23 f 0.02 in cyclohexane. By excitation in the second absorption band, at 215 nm, the fluorescence quantum yield is 4f,2= &f,l; therefore, only the /3 fraction (Table I) of the 'Bb states reaching the relaxed fluorescent state can be quenched and the quenching quantum yield beLe., 0.07 f 0.01 either in comes, at 215 nm, 4q,2= @4q,l, Me,Si or in cyclohexane. Photolysis Experiments. Nitrogen and oxygen were present in all indole samples, irradiated or not. The
The Journal of Physical Chemistry, Vol. 86, No. 22, 1982 4403
Photophysics of Indole in Tetramethylsilane
TABLE 111: Wavelength Effect upon the Quantum Yield of Nitrogen Formation Compared to the Quantum Yield for the Quenching of Indole by N,O at Atmospheric Pressure wavelength. , nm
Me,Si cyclo-
hexane
a
I$
quenching
300240
225210
0.19
0.07
0.23
0.07
e@,) 292
254
0.12 t 0.02
0.11 i 0.02 0.23a
214
0.07
*
0.02 0.06a
Reference 9.
quantities of each gas in blank experiments, not irradiated, were not constant as observed in ref 3 but varied from 1 to 25 mm3, probably due to minute air leakage and/or to a small decomposition of NzO. The ratio Nz/Oz in the blanks was always lower than its value in air, 3.7, with values ranging from 1.4 to 3.2. However, in irradiated samples the ratio Nz/Oz was always higher than 3.7 with values ranging from 7.1 to 14.7 whatever the irradiation wavelength. To obtain the photoproduced nitrogen we subtracted from the measured nitrogen the amount corresponding to the air nitrogen, i.e., 3.7 times the amount of oxygen in the sample. Such a procedure should give nitrogen values lower but certainly not higher than true values. In Table 111, the results obtained in Me4%at three wavelengths are reported together with the nitrogen formation in cyclohexane (from ref 9) and compared to the quenching quantum yields calculated above.
Discussion The drop of the fluorescence quantum yield of indole in Me4Si between the lLB,'Lb absorption bands and the 'Bb band whereas the fluorescence lifetime is constant implies that a deactivation process at the upper excited states, 'Bb, competes with the relaxation to the fluorescent state, both mechanisms being much faster than the radiative process. This deactivation process postulated in other solvents to be photochemical in nature and solvent-polarity dependent12was shown to correspond to the formation of solvated electron in aqueous solutionsgJ' and to the splitting of the N-H bond of indole in hydrocarbon solv e n t ~ .The ~ Stokes shift of indole, measured by its v, vf value in different solvents, was shown to correlatelo with Kosower's polarity parameter 2." From the v, - vf value found for indole in Me4% (0.13 pm-l) we can deduce a very low 2 value, i.e., a character of very low polarity for Me4% The drop of the fluorescence quantum yield is therefore expected to be due to the splitting of the N-H bond of indole in its 'Bb state rather than to the photoionization. However, the photoionization of indole was described to occur in Me4& from an energy threshold at 4.95 eV, i.e., 250 nm.3 Such a value, different from the 'Bb absorption band edge, implies according to ref 3 that the precursors for the photoionization reaction are the vibrationally excited singlet states of indole. In such a case, unless the photoionization quantum yield is too small to efficiently compete with deactivation to the 'Lb fluorescent state,the excitation spectrum should not follow the absorption spectrum. A continuous decrease in the fluorescence quantum yield with decreasing excitation wavelength should be expected as the electron quantum yield is claimed to show a continuous increase at wavelength lower than 250 nm.3 As shown in Figure 2 the drop of the fluorescence quantum yield corresponds to the overlap of the 'La, 'Lb (22) E. M. Kosomer, J. Am. Chem. SOC.,80, 3253 (1958).
Figure 5. Deactivation scheme for indole in Me,Si: (S,+) 'L, fluor'6, excited state, (F) fluorescence, (T) triplet state. escent state, (S2*)
absorption bands and the 'Bb band, but in each of these bands the fluorescence quantum yield is constant. Therefore, an energy-dependent photoionization in competition with the relaxation to the fluorescent state is unlikely unless the photoionization quantum yield is too low with respect to the accuracy of the present results to allow perceptible modification of the fluorescence excitation spectrum. According to ref 23 another possibility could be the repopulation of the fluorescent state from the recombination of the ion pair in the absence of an electric field or in the absence of electron scavenger. In this case, NzO, assumed to be an electron scavenger in Me4Si,3 should prevent the recombination and hence the excitation spectrum in the presence of NzO should be different from the excitation spectrum in the absence of NzO. This is not the case as shown by the superposition of the excitation spectra with and without NzO from 290 to 220 nm (Figure 3) although the photoionization was claimed to increase from 250 to 220 nm.3 In the wavelength range 250-220 nm, excitation spectra with and without NzO are superposed with an accuracy of f2%; hence, the photoionization quantum yield, if it increases from 250 to 220 nm,3 should remain lower than 2%. Furthermore, the present results contradict the assumption that NzO does not quench the indole fluorescence in Me4Si.3 The vibrationally excited 'La, 'Lb states are relaxed to the fluorescent state (Figure 5) before their quenching by NzO. Indeed, as the decrease in fluorescence lifetime parallels the decrease in fluorescence quantum yield (Table 11), the quenching is strictly dynamic in nature, i.e., results from a collision of NzO with indole in its first singlet excited state. The similarity of the Stern-Volmer curves obtained at 265 and 215 nm (Figure 4) shows that the 'Bb state is not quenched by NzO at the highest concentration used here. This implies a very short lifetime for the 'Bb excited state compared to the one of the fluorescent state, An important point to elucidate is whether this quenching could lead to nitrogen formation. Firstly, we have to discuss the following experimental contradiction: according to ref 3 the formation of Nz from NzO in Me4Si solution of indole was not found at 280 and 260 nm and then increased from 255 to 235 nm, whereas we find a rather constant nitrogen quantum yield at 292 and 254 nm and then a lower yield at 214 nm (Table 111). This disagreement cannot be explained by a biphotonic process due to the present laser excitation at 292 nm. Indeed, the 254-nm result was obtained with a continuous light source which delivered a dose 4-5 orders of magnitude lower than the dose obtained during the same time with the microsecond laser and leads nevertheless to the same nitrogen (23) E. Amouyal, A. Bemas, and D.Grand, Photochem. Photobiol., 29, 1071 (1979).
J. Phys. Chem. 1982, 86, 4404-4412
4404
formation yield as the 292-nm laser (Table 111). This nitrogen quantum yield is lower than the quenching quantum yield (Table 111) whereas in cyclohexane the nitrogen formation corresponds exactly to the quenchingg as this appears to be the case at 214 nm in Me4Si (Table 111). It seems then possible to conclude that the quenching reaction of indole by NzO can lead to nitrogen formation in Me4Si as in cyclohexane. A similar conclusion was obtained in a study of NzO effect on TMPD in cyclohexane.24 This implies that the use of N 2 0 as electron scavenger to determine the photoionization threshold of indole in Me,Si or cyclohexane is not suitable. The good agreement claimed for the photoionization threshold of indole in Me4Si measured by photoconductivity and electron scavenging by Nz03 seems then not at all convincing. In the light of the present results the photophysics of indole in Me& may be simply explained by the different relaxation of the excited states: only 36% of the ‘Bb excited states are relaxed to the fluorescent state while 100% of ‘La, ‘Lb excited states reach the relaxed fluorescent state as schematized in Figure 5. By analogy with its behavior in c y c l ~ h e x a n e ,we ~ , ~suggest that the remaining 64% of ‘Bb excited states of indole in Me& undergo N-H bond dissociation and that the deactivation ways of the fluorescent state are fluorescence for one half and mainly intersystem crossing to the triplet state for the other half as schematized in Figure 5. ~~~~
(24) J. T. Richards and J. K. Thomas, Trans. Faraday SOC.,66, 621 (1970).
Conclusion As in cyclohexane, indole in Me4Sisolution shows a drop of the fluorescence quantum yield between the ‘La, ‘Lb absorption bands and the IBb band. In degassed solution = 0.50 f 0.05 at wavelengths higher than 240 nm and @f = 0.18 f 0.02 a t wavelengths lower than 225 nm. The fluorescence lifetime, independent of the excitation wavelength, is 9.4 f 0.1 ns in degassed solution. As in cyclohexane, N 2 0 quenches the indole fluorescence in Me4Si. As shown by the parallel decreases of the fluorescence quantum yield and fluorescence lifetime, the quenching is dynamic in nature. It leads to the formation of nitrogen with a quantum yield higher in the first absorption band, 4(N2)= 0.12 f 0.02 at 292 or 254 nm, than in the second one @(N2)= 0.07 f 0.02 at 214 nm. The excitation spectra in the presence and in the absence of NzO are strictly superposed from 290 to 220 nm. The concentration effect of N 2 0 may be described by a Stern-Volmer relation with an identical slope by exciting both at 265 and 215 nm. The present data (NzO quenching, N2formation) do not confirm the previous experimental observations3 and therefore the photoionization threshold obtained with NzO as electron scavenger3 is questioned. Acknowledgment. We thank Professor IVI. Duquesne for continuous interest and stimulating discussions, Dr. J. Belloni for giving us full facilities to use a Carlo Erba gas chromatograph at the Laboratoire de Physico-Chimie des Rayonnements, Orsay, Dr. L. Chinsky and Dr. P. Y. Turpin for their help to perform irradiation experiments with the dye laser, and G. Tham for drawing figures. @f
Kinetics and Mechanism of Porphyrin-Photosensitized Reduction of Methylviologen M. Rougee,’ T. Ebbesen, F. Ghettl,+ and R. V. Bensasson Laboratoire de Biophysique, INSERM U.201, ERA 95 1 do C.N.R.S.,Museum National d’Histoire Naturelle, 75005 Paris, France (Received: March 22, 1982; In Final Form: July 7, 1982)
The electron transfer between cysteine (CysSH) and methylviologen (MVz+)is photosensitized by zinc tetrakis(psulfonatopheny1)porphyrin (ZnTPPTS) excited in its triplet state by visible light in aqueous solution at room temperature. Ground-state complexation of ZnTPPTS by MV2+gives a 1:l complex ( K = 1.5 X lo3 M-l) which does not yield any detectable transient upon excitation. The triplet state 3Pof the free ZnTPPTS (& = 0.9) is oxidized by MV2+( k = 6.9 X lo9 M-’ s-’) yielding P+. and MV+-with a cage escape yield of -0.5. Back-reaction of P+.and MV+. ( k = 1.3 X lo9 M-’ s-l) can be prevented by reaction of P+-in the presence of a large concentration of CysSH ( k = 1.7 X lo6 M-l s-l). CysS., the deprotonated form of the oxidized cysteine, reacts with CysS- to give the radical CysS-SCys-. which reduces MV2+( k = 8 X lo8 M-’ s-‘), so that the initial yield of MV+. is twice the product of the quantum yield of triplet formation of the photosensitizer and the escape yield of P+.and MV+.. A detailed mechanistic scheme of 11reactions has been deduced from the results of flash- and continuous-photolysis experiments.
Introduction The electron transfer A
+D
P hu
D+ + A -
(a)
between a donor D and an acceptor A photosensitized by a molecule P absorbing visible light is actively studied by several laboratories as a model reaction for photoconver‘Fellowship of the Consiglio Nazionale delle Ricerche (Italy). 0022-365418212086-4404$0 1.2510
sion and storage of solar energy. The most studied of these photoredox reactions use as acceptor methylviologen (MV2+,N,N’-dimethyL4,4’-bipyridinium cation), its reduced form the radical MV+. being able to reduce water and produce hydrogen via 2MV+. + 2Hz0
catalyst
2MV2++ Hz
+ 20H-
(b)
in the presence of a catalyst (Pt or hydrogenase). Such systems, so-called “Krasnovsky reactions”,’ were suggested 0 1982 American Chemical Society
