469
FLUORESCENCE QUENCHING IN POLAR SOLVENTS
The Electron-Transfer Mechanism of Fluorescence Quenching in Polar Solvents.
I. Dicyanobenzene as Quencher
by K. H. Grellmann, A.
R. Watkins, and A.
Weller*
Max Planck Institut f a r Biophysikalische Chemie, Karl Friedrich Bonhoeffer Institut, 0-3400Gotlingen, Germany (Receiued August 2, 1871) Publication costs assisted by M a x Planck Institut f 6 r Biophysikalische Chemie
The transient absorption spectra for the systems naphthalene, phenanthrene, coronene, pyrene, 1,2-benzanthracene, anthracene, perylene, tetracene with dicyanobenzene as quencher have been obtained (using acetonitrile as the solvent) by conventional flash photolytic methods. From these spectra it can be concluded that not only the ions but also the triplet states of the fluorescing hydrocarbons are intermediates in the quenching process. The question as to which mechanism could lead to the appearance of hydrocarbon triplets in the course of the quenching reaction is discussed.
Baurl and Weiss and Fischgold2 were the first authors to propose a fluorescence quenching mechanism proceeding by way of an electron transfer between fluorescer and quencher. I n this scheme the electronic energy of the excited molecule is dissipated, in the course of the electron transfer and reverse electron transfer steps, as vibrational energy to the solvent sink. One might expect, from this mechanism, a correlation between quenching efficiency and the oxidation-reduction properties of the molecules involved, and such a correlation has in fact been found for a considerable number of system^.^ Fluorescence quenching in nonpolar solvents is accompanied by the appearance of a new fluorescence emission a t longer wavelengths4 attributed to an excited charge transfer complex (heteroexcimer) which is formed from the initial collision complex (encounter complex)
'D*
+A
'D** * * * A+ '(D+A-)*
(1)
This supposition has been confirmed by four types of experimental observation. (i) The species emitting the fluorescence has a large dipole moment.5 (ii) The wavelength of the new emission shows a marked and very general correlation with the oxidation-reduction properties of the fluorescer and quencher? (iii) The absorption spectrum of the species emitting the new fluorescence shows a strong resemblance to the absorption spectra of the ions constituting the '(D+A-) complex.' (iv) The enthalpy of formation of the new species can be correlated with the oxidation-reduction properties of the fluorescer and quencher.8 As the polarity of the solvent increases, the fluorescence intensity and fluorescence lifetime of the heteroexcimer d e ~ r e a s e and , ~ in polar solvents the new fluorescence emission can no longer be detected. The quenching efficiency still depends on the oxidation and
reduction potentials (or ionization potential and electron affinity) of donor and acceptor,lO*llhowever, and (1) E. Baur, 2. Phys. Chem., Abt. B , 16, 465 (1932). (2) J. Weiss and H. Fischgold, 2.Phys. Chem. Abt. B , 32, 135 (1936). (3) A. Weller, Progr. React, Kinet., 1, 187 (1961); H. Leonhardt and A. Weller in "Luminescence of Organic and lnorganic Materials," Kallmann and Spruch, Ed., Wiley, New York, N. Y., 1962; A. Nakajima and H. Akamatu, Bull. Chem. Soc. Jap., 41, 1961 (1968); D. Schulte-Frohlinde and R. Pfefferkorn, Ber. Bunsenges. P h p . Chem., 72, 330 (1968); B. S. Solomon, C. Steel, and A. Weller, Chem. Commun., 927 (1969). (4) H. Leonhardt and A. Weller, Ber. Bunsenges. Phys. Chem., 67,791 (1963); N. Mataga, K. Ezumi, and K. Takahashi, 2. Phys. Chem. (Frankfurt am M a i n ) , 44,250 (1965); N. Mataga, T. Okada, and H. Oohari, Bull. Chem. Soc. Jap., 39,2563 (1966); H. Knibbe and A . Weller, 2. Phys. Chem., 56, 99 (1967); M. 8. Walker, T. W. Bednar, and R. Lumry, J . Chem. Phys., 45, 3455 (1966); N. Mataga and K. Ezumi, Bull. Chem. SOC.Jap., 40, 1355 (1967); N. Mataga, K. Ezumi, and T. Okada, Mol. Phys., 10, 201 (1966); W. R. Ware and H. P. Richter, J . Chem. Phys., 48, 1595 (1968) ; T. Okada, H. Matsui, H. Oohari, and H. Matsomoto, ibid., 49, 4717 (1968); D. Cros and P. Viallet, J . Chim. Phys., 67, 794 (1970); K. Mutai, Chem. Commun,, 1209 (1970); I. G. Lopp, R. W.Hendren, P. D. Wildes, and D. G. Whitten, J . dmer. Chem. Soc., 92, 6440 (1970). (5) H. Beens, H. Knibbe, and A. Weller, J. Chem. Phys., 47, 1183 (1967); M. G. Kuzmin and L. N. Guseva, Chem. Phys. Lett., 3, 71 (1969). (6) H. Knibbe, D. Rehm, and A. Weller, 2. Phys, Chem. (Frankfurt am Main), 56,95 (1967); E. A. Chandross and J. Ferguson, J . Chem. Phys., 47, 2557 (1967) : H. Beens and A . Weller, Acta Phys. Pol., 34, 693 (1968); D. Rehm and A. Weller, 2.Phys. Chem., 69, 183 (1970); M. A. F. Tavares, Trans. Faraday SOC.,66, 2431 (1970); D. Rehm, 2. Naturforsch. A , 25, 1442 (1970); D. W. Ellis, R. G. Haniel, and B. S. Solomon, Chem. Commun., 1697 (1970). (7) C. R. Goldschmidt and M. Ottolenghi, Chem. Phys. Lett., 4, 570 (1970); R. Potashnik and M. Ottolenghi, ibid., 6, 525 (1970); H. Masuhara and N. Mataga, ibid., 6, 608 (1970). (8) H. Knibbe, D. Rehm, and A. Weller, Ber. Bunsenges. Phys. Chem., 73, 839 (1969). (9) N. Mataga, T. Okada, and N. Yamamoto, Bull. Chem. SOC. Jap., 39, 2562 (1966); N. Mataga, T. Okada, and K. Ezumi, Mol. Phys., 10, 203 (1966) ; H. Knibbe, K. Rollig, F. P. Schafer, and A. Weller, J . Chem. Phys., 47, 1184 (1967); N. Mataga, T . Okada, and N. Yamamoto, Chem. Phys. Lett., 1, 119 (1967); K. Kaneta and M. Koisumi, Bull. Chem. SOC.Jap., 40, 2254 (1967); S. Murata, H. Kokubun, and M. Koizumi, 2. Phys. Chem. (Frankfurt am Main), 70, 47 (1970). (10) K. Kaneta and M.Koisumi, Bull. Chem. Soc. Jap., 40, 2254 (1967); D. Rehm and A. Weller, Ber. Bunsenges. Phys. Chem., 73, 834 (1969).
The Journal. of Physical Chemistry, Vol. 76,No. 4, 1978
470
K. H. GRELLMANN, A. R. WATKINS,AND A. WELLER
fluorescence quenching still seems to take place by electron transfer, leading, this time, to solvated ions 'D*
+A
'D*.
.
I
I
I
I - 0.01
. A -+ De+ + A,-
(2) Flash s t u d i e ~ l ~ have - ' ~ shown that in fact ions are produced in quenched systems in acetonitrile and dimethylformamide; however, triplet states are also observed, which, at the high quencher concentrations used, cannot be formed by intersystem crossing from the excited molecule. The present investigation was made in order to gain more information about what part these intermediates play in the quenching process. 9
r i r
e DCB
0 5
I
NwhlhPUle 0.91
x IOJM
0,091
M
i
Experimental Section The following compounds were recrystallized and zone-refined: pyrene, l,Zbenzanthracene, perylene, anthracene, and naphthalene. Coronene was recrystallized and zone-sublimed. Phenanthrene was freed from anthracene by the method of Bachmann116 recrystallized twice from ethanol, chromatographed on an A1203 column and zone-refined. In subsequent experiments with phenanthrene purified in this way no trace of anthracene could be detected. Tetracene was purified by recrystallization and sublimation; commercial p-dicyanobenzene (hereafter DCB) was recrystallized, sublimed and then zone-melted under 3-4 atm of nitrogen. The solvent used in most experiments, acetonitrile, was prepared from uv grade acetonitrile (Rilerck UVASOL) by refluxing over Pz05 followed by fractional distillation. This product proved to be 100% transparent to 200 nm. Flash experiments were carried out with an apparatus having a stored electrical energy of 150 J available for the flash, which had a half-life of 5 psec. A photoelectric recording system with a 1P28 photomultiplier and a Tektronix 549 Oscilloscope was used. The transient spectra were obtained by varying the monochromator wavelength of the detection system through the entire wavelength range (300-730 nm) at 5-nm intervals, flashing a t each of these wavelengths and recording the transient absorption immediately after the flash had ceased (about 25 psec after firing). The maximum sensitivity of the transient fibsorption measurements was about 0,001 optical density units, the detection system becoming less sensitive (as a result of the photomultiplier and monitoring light source characteristics) at the extremes of the wavelength range. The flash cell used consisted of a cylindrical quartz cell, 10 cm long, to which two reservoirs mere attached. Using these reservoirs, the solutions to be investigated (fluorescer and quencher) were degassed separately by repeated freeze-pump-thaw cycles, and then mixed prior to flashing, in order to avoid the occurrence of any light induced reaction between the two in the presence of air. A 0.1 M solution of DCB in 1-cm filter cells inserted between the flash lamps and the cell ensured that only the donor molecule (in all cases the The Journal of Physical Chemistry, Vol. 76, N o . 4- 1978
Figure 1. The transient absorption spectrum (upper half) and corresponding literature spectra (lower half) for the system naphthalene-DCB. N + spectrum from ref 16, DCB- from ref 17, aN*from ref 18.
Y
kK
Figure 2, The transient absorption spectrum (upper half) and corresponding literature spectra (lower half) for the system phenanthrene-DCB. Phf spectrum from ref 16, 3Ph* from ref 18.
aromatic hydrocarbon) was excited by the flash light. In no case was there any evidence of irreversible photochemical changes occurring during the flash experiments, and no complex formation in the ground state between fluorescer and DCB could be detected. Fluorescence quenching measurements were carried out with a Perkin-Elmer Fluorescence Spectrophotometer R/IPFBA in conjunction with a Phillips Digital Multimeter PM1412, b y means of which the observed (11) D. Rehm and A. Weller, Israel J. Chem., 8, 259 (1970). (12) H. Leonhardt and A. Weller, 2. Phys. Chem. (Frankfurt am Main), 29, 277 (1961); Ber. Bunsenges. Phys. Chem., 67, 791 (1963). (13) H. Knibbe, D. Rehm, and A. Weller, B e y . Bunsenoes. Phys. Chem., 72, 257 (1968). (14) K. H. Grellmann, A. R. Watkins, and A. Weller, J . Lumkz., 1 , 2 , 678 (1970). (15) W. E. Bachmann, J . Amer. Chem. SOC.,57, 555 (1935).
471
FLUORESCENCE QUENCHINGIN POLAR SOLVENTS
vhK
Figure 3. The transient absorption spectrum (upper half) and corresponding literature spectra (lower half) for the system coronene-DCB. C f spectrum from ref 19, ,C* from ref 18.
I I
Figure 5 . The transient absorption spectrum (upper half) and corresponding literature spectra (lower half) for the system 1,2-benzanthracene-DCB. 3BA* spectrum from ref 21.
1 05
Y
hK
Figure 4. The transient absorption spectra (upper half) and corresponding literature spectra (lower half) of the system pyrene-DCB in ethanol and acetonitrile. P+ spectrum from ref 20, 3P* from ref 21.
fluorescence intensity a t a fixed wavelength could be read off accurately. A solution of the fluorescing substance without added quencher served as a standard, by means of which fluctuations in the exciting light intensity could be corrected for. Both the flash and fluorescence quenching experiments were carried out at room temperature.
Results and Discussion The transient absorption spectra obtained in the flash experiments described above, together with the relevant spectra from the literature, are presented in Figures l-8.18-23 Extinction coefficients for triplet states were taken from ref 24. Extinction coefficients for the following species were not available in the literature: the triplet states of perylene and tetracene, DCB radical anion, and the radical cations of coronene, phenanthrene and naphthalene. The corresponding
Figure 6. The transient absorption spectrum (upper half) and corresponding literature spectra (lower half) for the system anthracene-DCB. A + spectrum from ref 20, 3A* from ref 18.
spectra have therefore been drawn in arbitrary units. Comparison of the flash spectra with the literature spectra shows that the radical cation and the triplet state of the excited molecule (in these systems the elec(16) T. Shida and W. H. Hamill, J . Chem. Phys., 44, 2375 (1966). (17) A. Ishitani and 5. Nagakura, Theor. Chim. Acta, 4, 236 (1966). (18) G . Porter and M. W. Windsor, Proc. Roy. SOC.Ser. A , 245, 238 (1958). (19) P. Bennema, G. J. Hoijtink, J. H. Lupinski, L. J. Oosterhoff, P. Selier, and J. D. W. van Voorst, Mol. Phys., 2 , 431 (1959). (20) W. I. Aalbersberg, G . J. Hoijtink, E. L. Maokor, and W. P. Weijland, J . Chem. Soc., 3049 (1959). (21) W. Heinzelmann and H. Labhart, Chem. Phys. Lett., 4, 20 (1969). (22) "DMS UV Atlas of Organic Compounds," Butterworths, London, 1966. (23) H. Staerk, private communication. (24) J. S. Brinen, in "Molecular Luminescence," W. A. Benjamin, New York, N. Y., 1969, p 333.
The Journal of Physical Chemistry, Vol. 76,No. 4, 1972
K. H. GRELLMANN, A. R. WATKINS,AND A. WELLER
472
1' i
I
I
Figure 9. Schematic representation of the energy levels of the intermediates involved in the quenching process. Y LK
Figure 7. The transient absorption spectrum (upper half) and corresponding literature spectra (lower half) for the system perylene-DCB. Pe+ spectrum from ref 22, *Pe*from ref 23.
Figure 8. The transient absorption spectrum (upper half) and corresponding literature spectra (lower half) for the system tetracene-DCB. T+spectrum from ref 22, aT*from ref 18.
tron donor) together with the radical anion of DCB, are formed in the systems phenanthrene-DCB in acetonitrile, pyrene-DCB in ethanol, and anthraceneDCB and perylene-DCB in acetonitrile. The spectrum of the radical cation of 1,Bbenzanthracene has to date not been reported, and the unfavorable position of the ground state absorption of 1,2-benzanthracene prevented identification of the DCB- absorption a t 29 kK in this system; however, the riplet spectrum can be clearly identified in the experimentally obtained flash spectrum. The system coronene-DCB shows a peak a t 21 kK which is probably due to both the coronene triplet and the coronene radical cation; the peak between 28 and 29 kK is probably made up of the coronene radical cation and DCB-. No explanation has yet been found for the peaks a t 25 and 16 kK. The transient absorption spectrum of the system naphthalene-DCB is made up of the DCB- absorption a t 29 The Journal of Physical Chemistry, Vol. 76,No. 4, 19W
kK and a weaker absorption centered at 24 kK, presumably due to the triplet and radical cation of naphthalene, although the published spectrum of the latter species extends to only 23 kK, making a definite assignment impossible, In no case can it be concluded from the observed transient spectra that the triplet of the fluorescing molecule is absent; it is either clearly recognizable or overlapped by the absorption of some other species. Figure 9 shows the energies, calculated from published spectral and electrochemical data, of the singlet, triplet, and ionic states involved in the quenching reaction. It can be seen that in all cases the quenching reaction leading to the free ions is thermodynamically favored, although only slightly in the case of coroneneDCB. The favorable free energy changes are reflected in the quenching data given in Table I, where the measured quenching constants K are given. The extent to which the donor fluorescence was quenched in the systems investigated in the flash experiments is also included. Also included in Table I are the quantum yields of the donor triplet (&) and of the donor cation (&). The quantum yield of DCB-, +A, was assumed to be equal to &. These quantum yields were obtained by flashing the donor molecule twice under the same conditions, once with acceptor and once without. The quantum efficiency of transient formation in the quenched system is given by
A To
E.T+T~
d , = --
E.
(3)
where ET and E. are the extinction coefficients, at the wavelengths of maximum absorption, of the donor triplet and of the transient, A and To are the observed maximum optical densities (at these wavelengths) of the transient formed in the presence of acceptor, and of the donor triplet in the absence of acceptor, respectively, and + T ~is the quantum efficiency of triplet formation for the donor when no acceptor is present. This method presupposes well-separated spectra of the transient species involved, and a knowledge of the
473
FLUORESCENCE QUENCHING IN POLAR SOLVENTS releva,nt extinction coefficients and of the quantum yield of donor triplet formation in the absence of acceptor. I n fact, these requirements were difficult to meet, and in most cases only the quantum yield of triplet formation in the quenched system could be obtained.
within the lifetime of the flash. This quenching reaction may be represented as
1D*
Donor
Naphthalene Phenanthrene Coronene Pyrene 1,2-Benzanthracene Anthracene Per ylene Tetracene
K ,M - 1
175" 2800 280" 5000a 590a 77.3 93.3 30.5
94.170 99.6% 93.3% 99.6% 98.2% 78.670 81.6% 59.170
hb
... ... ...
0.46 0.45 0.69 0.05
0.51
...
ID*. . . .A
3D* + A
0.39 0.53
...
kI kz
Table I: Data on the Quenching of Fluorescence and the Transient Quantum Yields for the Systems Studied Extent of quenching at quencher ooncn used
+A
... ...
0.05"
...
a From ref 11. The intersystem crossing quantum yields used in calculating these quantities were taken from the collection in J. B. Birks, "Photophysics of Aromatic Molecules," Wiley-Interscience, New York, N. Y., 1970. Calculated from eq 3 using the value of the perylene triplet-triplet extinction coefficient reported by R. Bensasson and E. J. Land, Trans. Faraday Xoc., 67, 1904 (1971).
Comparison of the data of Figures 1-9 and of Table
I leads to some interesting conclusions. Firstly, the flash data confirm the generality of the charge transfer mechanism of fluorescence quenching, since in nearly all cases radical ions of the fluorescer and/or quencher could be identified in the flash spectra. Secondly, the triplet, in most cases, also appears to be an intermediate in the quenching reaction. Since quenching is substantially complete in most cases, intersystem crossing from unquenched donor molecules cannot account for the appearance of the donor triplets in the flash spectra, with the possible exception of the systems anthracene-DCB, perylene-DCB, and tetracene-DCB. It is difficult to assess the origin of the triplets observed in these experiments. In the cases investigated here the triplet level of the donor lies below the energy level of the free ions (which is 0.06 eV higher than the corresponding ion pair energy) and a process, whereby triplets are produced from a recombination reaction of the free radical ions produced by the initial flash,ll is clearly energetically possible. It has been established from chemiluminescence experiments25 that ion recombination can lead to triplet formation, and the diffusion controlled nature of this process Could, with the ion concentrations produced in these flash experiments, lead to a substantial formation of donor triplets
where formation of the ion pair (k3) is decisive in determining the energy gain and, therefore, the rate constant, of the overall quenching reaction. However, other possible explanations of triplet formation, that the donor triplets are formed from the encounter complex lD*. .A,26or from the solvated ion pair D,+. AB-,l 3 cannot be discounted. The common intermediate in both cases could be a contact ion pair (D+, A-) which differs from the heteroexcimer insofar as the doublet character of the radical ions is still preserved, and from the solvated ion pair by having no solvent molecules in between the ions. It is not at all impossible that the charge transfer interaction in this ion pair results in such strong spin-orbit coupling that the correspondingly enhanced intersystem crossing rate can effectively compete with the dissociation rate of the contact ion pair. The triplet quantum yields reportedjn Table I are complex quantities involving a number of steps; for example, in the reaction scheme 4 above, the steps corresponding to radical ion formation from the excited molecule ID*, and to triplet formation from the radical ions. It is remarkable, however, that, in the case of pyrene, such a high proportion of excited molecules are converted into triplets and radical cations, only 3% of the originally excited molecules reaching the ground state directly, whereas, in the case of perylene, this figure reaches 90%.27 a
+
-
a
a
Acknowledgments. The authors are grateful for the assistance of Ii!Irs. S. Reiche, who carried out much of the experimental work, and to Dr. B. Nickel, who provided considerable help with the zone-refining of the chemicals used. A. R. W. wishes to thank the Alexander von Humboldt-Stiftung for the award of a fellowship, during the tenure of which this work was carried out. Thanks are also due to Dr. W. Kuhnle, who purified the solvents used in this work. (25) A. Weller and K. Zachariasse, J . Chem. Phys., 46, 4984 (1967). (26) C.R. Goldschmidt, R. Potashnik, and M. Ottolenghi, J . P h y ~ . Chem., 75, 1025 (1971). (27) NOTEADDED IN PROOF.An analysis of the kinetics of formation and decay of both ions and triplets shows that the initial increase in triplet concentration expected from the reaction scheme 4 will occur in B time too short to be observed using the flash apparatus described here.2a
The Journal of Physical Chemistry, Vol. 76,No. 4, 1974
