J, CHEM. SOC. FARADAY TRANS., 1993, 89(24), 4271-4273 4271 Comparative Electron-transfer Quenching Rates of 9,IO-Dicyanoanthracene by Various Donors in Solvents of Different Polarities N. Ghoneim, C. Hammer, E. Haselbach, D. Pilloud and P. Suppan Institute of Physical Chemistry, University of Fribourg , Switzerland P. Jacques Laboratoire de Photochimie Generale ENSC, Mulhouse , France Multiple Rehm-WeIler plots of the rate constants of quenching of 9,lO-dicyanoanthracene fluorescence by elec- tron transfer with a variety of donors of different adiabatic ionization potentials have been observed in polar solvents (1-fluoropentane and acetonitrile). The donors fall into two classes which can be described as ‘n’ (localised positive charge in the cation) and ‘n’(delocalised charge), the former being the more efficient quen- chers. Possible reasons for this difference are discussed : electrostatic and solvation energies, internal reorgan- isation energy.Current evidence favours the former explanation. In a recent study of the quenching of the fluorescence of 9,lO-dicyanoanthracene (DCA) by various electron donors in hexane it was concluded that donors could be separated in two classes of different quenching efficiencies.’ When the quenching rate constants k, are plotted against the adiabatic ionization potentials E,, ,of the donors, two distinct Rehm- Weller plots are obtained with a separation of some 1 eV. The two classes of donors are described as ‘n’ and ‘n’,follow-ing the distribution of the positive charge in the radical cation formed by electron transfer (et): in n-donors like monoamines the charge is strongly localised in the n-orbital of the N-atom, whereas it is delocalized over the n orbitals of n-donors such as aromatic molecules.Here we investigate why this difference in charge localisation should influence the rate constant of et. One possible explanation resides in the Coulomb term of the Gibbs energy change in eqn. (1) or (2) for this process:’ or AG = Ei,,(D) -E,,(A) -E-1” -E* + C (2) In the latter the redox potentials E,,, Ercdare replaced by the of the donor and the acceptor’s gas-phase electron affin- ity E,, with the addition of the total solvation energy EWlvof the ion pair. E* is the energy of the excited state involved in et, and C is the electrostatic interaction energy (Coulomb term) gained by bringing the product ions from infinity to the actual distance of et, in general the distance of Van der Waals contact.In the classical point charge model this term is C=-44’ (3)EU for product molecules of electrostatic charges q, q’ at an encounter distance a (centre-to-centre) in a solvent of static relative permittivity E. In a more realistic model at the molec- ular scale, a distinction should be made between the centre- to-centre distance a and an eflectiue average charge-to-charge distance p. Fig. 1 shows a sketch of two localized ions (a)and a localized-delocalized ion pair (b). In the first case the effec- tive average centre-to-centre distance p is equivalent to the charge-to-charge distance a, but in the second case the Coulomb term is the sum of i Coulomb terms with variable charge separations a, 2 a, c=-c--44’4 4: E a, Ep (4) for point charges 4:at distances a, from the point charge q.Even this simple model shows that for a fixed value of a the Coulomb term must be smaller when one of the ions presents a delocalised charge distribution. In the case of a highly symmetrical arrangement of the ion pair with equal charges qfin Fig. l(b), (e.g. in C,HZ) all the charge-to-charge distances a,; eqn. (4) then leads to In principle, the different charge distribution in n- and n-donors will yield different C terms. However, in a non-polar solvent like hexane the variation of C with a is large, and the difference between the n- and n-donors could arise also from different values of a at which et occurs in the two classes..The question then arises of whether this classification of the quenchers in the two classes is still observed in solvents of higher polarity where C and therefore its variation with a is smaller. We report here the results of quenching experiments of the fluorescence of ‘DCA* in 1-fluoropentane (1FP) and acetonitrile (MeCN). The choice of 1FP as a solvent of inter- mediate polarity [Onsager polarity function f(D)= 0.621 was dictated by several considerations : Its high ionization poten- tial preventing its action as a quencher of ‘DCA*, and its good solubility properties towards the donors and acceptors used in this study.(a1 (b 1 Fig. 1 Models of (a) localised and (b) delocalised charge distribu- tions for the calculation of the electrostatic interaction energy 4272 J. CHEM. SOC. FARADAY TRANS., 1993, VOL. 89 Experimental t The fluorimeter used for this work and the experimental pro- cedures have been described in an earlier publication, as have the samples and the methods of purification.' The fluorescence lifetime of 'DCA* in 1FP could not be found in the literature and was therefore measured on our laser flash photolysis apparatus3 as 14 ns. This value is used for the calculation of the quenching rate constants from Stern-Volmer plots which were linear in all cases reported here. Exciplex luminescence was less of a problem in these polar solvents than in hexane (cf:ref.1). 1FP was obtained I lofrom Aldrich and MeCN from Fluka (UV grade). The sol- vents were used without further purification. 6.5' + 7.0 8.0 9.0 10.0 Results Ei,a/ev Fig. 3 Rehm-Weller plots for the quenching of the fluorescence of Table 1 shows the observed quenching rate constants with a 'DCA* in l-fluoropentane. Symbols as in Fig. 2. range of donors, together with their Ei, values and their clas- sification as 'n-' or 'K-' donor^.^ The Rehm-Weller plots of Fig. 2-4 show the quenching behaviour in hexane, 1FP and but the points are renumbered for ease of comparison with MeCN, respectively. The plot of Fig. 2 is taken from ref. 1 Fig. 3 and 4. The appearance of two well separated Rehm-Weller plots in 1FP and MeCN is evident.The separation decreases, A 12 11.0110.0 10.0 9.0 8.0 7.0 ' 7.0 8.0 9.0 10.0 6.0 b 7.0 8.0 9.0 10.0El. alev Ei,a/evFig. 2 Rehm-Weller plots for the quenching of the fluorescence of 'DCA* in hexane. a,n-donors; 0,n-donors. (The donors' reference Fig. 4 Rehm-Weller plots for the quenching of the fluorescence of numbers correspond to Table 1.) 'DCA* in acetonitrile. Symbolsas in Fig. 2. Table 1 Adiabatic ionization potentials of donors and quenching rate constants for the electron transfer quenching of the fluorescence of 9,lO-dicyanoanthracene in different solvents log (k,/dm3 mol-' s-') donors 4,JeV" hexanec l-fluoropentane acetonit rile ndonors 1 anthracene 7.45 10.36 10.40 10.20 2 hexamethylbenzene 7.85 10.15 10.07 10.23 3 durene 8.04 9.56 9.92 10.15 4 mesi t y lene 8.41 8.28 8.0 8-70 5 p-x y lene 8.44 7.84 7.53 8.79 6 toluene 8.82 7.51 7.33 7.08 7 ethylbenzene 8.77 7.06 ndonors 8 DABCO~ 7.20 10.30 10.33 9 diethylamine 8.01 10.25 10 benzylamine 8.70 9.76 10.13 10.00 11 n-butylamine 8.71 9.90 12 2-methoxyethylamine 8.90 9.80 13 3-chlorop yridine 9.10 8.09 8.51 8.98 14 pyridine 9.25 9.19 9.43 9.53 15 1,24imethoxyethane 9.30 7.72 7.12 6.72 From ref.6. DABCO = 1,4-diazabicyclo[2.2.2]octane. From ref. 1. J. CHEM. SOC. FARADAY TRANS., 1993, VOL. 89 Table 2 Coulomb and solvation terms (in eV) in different solvents ~ solvent AC('n' -'n') ~ AE,Jn' -'n') hexane 0.9 0.1 1FP 0.6 0.15 MeCN 0.45 0.2 however, to 0.6-0.7 eV (within the scatter of the experimental points) as compared to ca.1eV in hexane.' Discussion and Conclusion The classical point-charge Coulomb term [eqn. (3)] clearly cannot account for these observations: With E = 37 for MeCN, it would predict that a single Rehm-Weller plot should be obtained in such a highly polar solvent, regardless of possible moderate variations of a between the two classes. The modified Coulomb term [eqn. (6)] which implies no screening effect of the solvent in the case of ions in contact predicts on the other hand the values5 e2c=--for a non-polar solvent (E = 2) (6a)2a The latter is much larger than the value calculated from eqn.(3) and about half of that calculated for hexane [eqn. (6a)l. According to this effect alone, the separation of the Rehm- Weller plots of n- and n-donors should decrease from ca. lev observed in hexane' to around 0.5 eV in MeCN, in agree- ment with the experimental results. Note that in the comparisons of Fig. 2-4 the abcissae are the gas-phase Ei,a values and that both C and Emlv are omitted. The Born equation [eqn. (7)] gives Esolvfor an ion of charge q and radius r as5 Emlv= -4"2r (1 -:) (7) The difference between the n- and n-donors resides in the values of I, i.e. rn < I,. Then so that for a fmed value of the term (l/an -l/aJ the differ- ence AE,,lv increases with solvent polarity. This favours the n- over the n-donors as does the Coulomb term.The effect of solvation is therefore to increase the separation of the Rehm- Weller plots between the n- and z-donors, which further- more increases with solvent polarity. The values of the solvent polarity function (1 -1/~)are 0.47 for hexane, 0.71 for 1FP and 0.97 for MeCN. Hence, differences in solvation energies alone cannot explain the observed separations of the Rehm-Weller plots. In fact MeCN is the most polar solvent in this series, yet it gives the smallest separation. The most likely explanation is a combination of the contact Coulomb term [eqn. (6)] and of the solvation term [eqn. (S)] as given in Table 2. The separation of the Rehm-Weller plots of n- and n-donors is then attributed to the sums of these energy terms = 1.0 eV for hexane, 0.75 eV for 1FP and 0.65 eV for MeCN.The Coulomb term decreases with solvent polarity, while the solvation term increases, as expected. Apart of these solvent related effects, a purely molecular origin of the difference between n- and z-donors could be envi~aged.~The internal reorganization energy is greater for amines (n-donors) than for aromatics (n-donors) because of the change in geometry on going from the pyramidal struc- ture around the N-atom in the neutral to the planar one in the cation. This effect, however, is not in agreement with the observation that n-donors are more efficient quenchers than n-donors of the same The role of internal structural reorganization, therefore, cannot play a decisive role. It appears that the solvent effect on the ion pair, both on its solvation and its internal electrostatic interaction, must play the decisive role since the separation of the Rehm-Weller plots clearly decreases with solvent polarity. This work is part of project No.20-34071.92 of the Schwei- zerischer. Nationalfonds zur Forderung der Wissenschaftli- chen Forschung. References 1 P. Jacques, E.Haselbach, A. Henseler, D. Pilloud and P. Suppan, J. Chem. SOC.,Faraday Trans., 1991,87,3811. 2 D. Rehm and A. Weller, Zsr. J. Chem., 1970,8,259. 3 D. Noukakis and P. Suppan, J. Lumin. 1991,47,285. 4 S. G. Lias, J. E. Bartness, J. F. Liebman, J. L. Holmes, R. D. Levin and W. G. Mallard, J. Phys. Chem. Ref: Data, 1988, 17, Suppl. 1. 5 P. Suppan, J. Chem. SOC.,Faraday Trans. 1,1986,82,509. 6 M. Born, 2. Phys., 1920,1,45. 7 F. A. Carroll, M. T. McCall and G. S. Hammond, J. Am. Chem. SOC.,1973,95, 315. Paper 3/03307D; Received 9th June, 1993