J. CHEM. SOC. FARADAY TRANS., 1994, 90(2), 279-285 Effects of Protolytic Interactions on the Photophysics of Phenyl Pyridyl Ketones Fausto Elisei, Gianna Favaro" and Fausto Ortica Dipartimento di Chimica Universita di Perugia , I46123 Perugia Italy The photophysics of phenyl 3-pyridyl ketone (3-PPK) and phenyl 4-pyridyl ketone (CPPK) were investigated in aqueous solution as a function of pH (0-10) by steady-state and pulsed emission spectroscopy and by nanosec- ond laser flash photolysis. From phosphorimetric and triplet-triplet absorption acid-base titrations of the excited state, two protonation steps were evidenced for 3-PPK, which involved both the carbonyl oxygen and nitrogen atoms. The triplets of the neutral molecule and of the mono- and di-cation were identified from the shape of the transient absorption spectra, analysis of the kinetic profiles and effect of charged quenchers.For 4-PPK, the titration curves showed only one inflection point. Phosphorescence emission and T, +T,, absorption spectra, over the whole pH range explored, showed only the triplet of the neutral molecule. Indirect evidence of protonation was obtained from the weaker emission and absorption intensities and shorter lifetimes with increasing acidity. The comparison with ground-state pKs showed that the basicity of these molecules increases greatly upon electronic excitation. Interest in pyridyl ketones as potential water-soluble photo- sensitizers, which could possibly be used also in acidic solu- tions, stimulated this study of the effects of pH on their pho tophysics.Previous studies on the photo~hemistry'-~ and photo physic^^*^ of phenyl pyridyl ketones (n-PPKs) gave information on the nature and reactivity of their excited states. These molecules were found to possess lowest singlet and triplet states of n,n* character with excitation energies slightly lower than that of benzophenone owing to the electron- withdrawing effect of the heterocyclic nitrogen. In aqueous solutions, only two of the three isomers, the 3- and 4-PPKs, exhibited a triplet behaviour which was similar to that of benzophenone (B). In contrast, the 2-isomer, which still showed similar absorption spectra and low-temperature n,n* triplet emi~sion,~ underwent photochemical rearrangement with formation of coloured photo product^,^ when irradiated in a water solution.In such a medium, this isomer did not show either room temperature phosphorescene, or transients which could be securely attributed to the n,n* carbonyl triplet, at least with nanosecond time resolution. This paper is concerned with the 3- and 4-PPKs, which were found to be photostable enough to be used as triplet photosensitizers in benzene, as well as in water solution. The protolytic interaction of the 3-PPK in the excited state was previously investigated by phosphorescence quenching and sensitization measurements. The results showed inter- esting pH effects on phosphorescence emission intensity and triplet lifetime. It seems worthwhile to re-examine these mol- ecules by using nanosecond laser flash photolysis and a high- sensitivity spectrofluorimeter in order to gain more insight into the protolytic interaction in the excited state and the transients produced under laser excitation. The ionization constant of an electronically excited mol- ecule may differ from that of the ground state by several orders of magnitude due to the different electronic distribu- tion in the excited state.In PPKs, it is difficult to establish the sequence of the excited-state protonations because of the presence of two basic centres (on the carbonyl oxygen and on the heterocyclic nitrogen). By combining kinetic and steady- state phosphorescence measurements with laser flash pho- tolysis measurements on transient absorption intensity and decay, a reliable scheme is proposed for protolytic interaction of triplet excited 3- and 4-PPK in the 0-10 pH range.Experimental Materials The 3- and 4-PPKs were purchased commercially (Aldrich). The 3-PPK (mp 38°C) was recrystallized from ethyl ether at -25 "C and 4-PPK (mp 75 "C) from water-ethanol. Britton buffer solutions were used from pH 10 to pH 2, at constant ionic strength (p = 0.01); HClO, solutions were used in the higher acidity range, down to pH = 0. The pH values of the solutions were determined with an Orion digital pH-meter SA-520 equipped with an Orion 9103 semimicro electrode. Equipment The absorption spectra were recorded on a Perkin-Elmer Lambda 5 (double beam) spectrophotometer.The phosphorescence measurements were performed on a Spex Fluorolog-2 FL112 spectrofluorimeter controlled by the Spex DM3000F spectroscopy computer. A home-made apparatus, previously described,6 was used for the phosphorescence decay time measurements in fluid solution (order of magnitude: 1-300 p).The decay curves were mono-exponential over at least two or three half-lives. The reproducibility of the mean lifetimes was within 20%. For laser flash photolysis measurements, the 308 nm line from an Xe-HCl excimer laser (Lambda Physik LPX 105) or the 347 nm line from a ruby laser (J.K., second harmonics) were used. The laser energy was less than 10 mJ per pulse. Details of the equipment and data processing methods have been described elsewhere.' The transient spectra were obtained by a point-to-point technique over the spectral range 300-700 nm.Phosphorescence decay kinetics were also recorded with the same apparatus to compare the decay rate constants of absorption and emission under the same experi- mental conditions. Measurement Conditions Phosphorescence quantum yields were determined in dilute (c < 3 x lo-, mol dm-3), low absorbance (A d 0.07) solu-tions, using quinine bisulfate in 0.5 mol dmP3 H,S04 as a standard. Intersystem crossing yields were obtained from measurements of sensitized biacetyl phosphorescence, using benzophenone ($Isc = 1) as a reference molecule. For the phosphorimetric titrations, the solutions were excited at an isosbestic point between the neutral and nitrogen-protonated form.For the triplet-triplet absorption titrations the absorbance of the solutions was adjusted on a constant value (A = 0.25) at the exciting wavelength : the transient absorptions were monitored at the visible maximum wavelength, 50 ns after the laser pulse. Transient lifetimes were measured as a function of pH keeping the sample concentration constant (c = 7 x mol dm-3, with 308 nm excitation, and 2 x mol dm-3, with 347 nm excitation). The expected accuracy in lifetime was within 20%. Quenching experiments with the inorganic anion Cr(CN); -were carried out adjusting the quencher concentra- tion in the range 1.5 x 10-2-1.5x mol dm-3, depend- ing on the lifetime of the ketone.Fresh solutions of salt were used in order to avoid thermal aquation. At the exciting wavelength (347 nm) photoaquation was negligible owing to the low absorbance of the anion. All measurements were performed at room temperature. The solutions were de-aerated with oxygen-free nitrogen or argon, unless otherwise indicated. Results Photophysical Parameters based on Room-temperature Phosphorescence Emission The 3-and 4-PPKs were previously found to exhibit phos- phorescence emission in aqueous neutral solutions, like ben- zophenone. The typical vibrational structure, showing the carbonylic progression, and rather long lifetime (in the micro- second domain) indicated that the lowest triplets are n,n* in ~haracter.~ Some photophysical parameters of 3-and 4-PPK in aqueous solution are compared with those of benzophenone in Table 1.Triplet energies were taken from the 0-0 transition of the low-temperature (77 K) phosphorescence ~pectra.~ Phosphor-escence yields and lifetimes were extrapolated at infinite dilu- tion in order to compare their values independently from concentration. The self-quenching rate parameters of PPKs, k,, were obtained as the slope of a plot of the inverse of the experimental phosphorescence lifetime, l/~~,~,against con- centration (2 x 10-'-2 x mol dm-3), keeping the pH of the solution constant. The intercept of the plot T-' us. con-centration gives the sum of the radiative, k,, and non-radiative, k,, , rate parameters. The radiative constant was determined through the relationship: k, = (&c 7)-'4,. The values obtained are in reasonable agreement with those cal- Table l Triplet-state parameters of 3-PPK and CPPK obtained from phosphorescence measurements in alkaline aqueous solution at room temperature, compared with those of benzophenone (E, refers to 0-0 transition from the low-temperature phosphorescence spectra) 3-PPK 4-PP benzophenone E,"/kJ mol-' 4P 'TIPs 278 1.3 x lo-' 70' 273 2.6 x 15 280 4.4 x 200 lo-' k,/dm34Isc mol-'s-' 6 x 0.7 4.4 x 0.5 107 1.7 x 1 loac kp/s-k,,ls - 2.6 x 10' 1.4x lo4 3.5 x 10' 6.6 x lo4 2.2 x 10' 4.8 x lo3 Data from ref.4; from ref. 5; from ref. 18. J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 culated from the phosphorescence lifetime measured in a glassy solution at 77 K (kp = l/t = 2.8 x lo2 sC1 and 5.9 x lo2 s-l, for 3-and 4-PPK, respectively).* Both phosphorescence intensity and lifetime decreased, as the acidity of the solution increased, down to undetectable values at pH < 0 for 3-PPK and pH < 4 for 4-PPK.In a previous paper,' where we studied the effect of pH on the phosphorescence of 3-PPK, no change in the spectral dis- tribution of the emission could be detected in the pH range 2-10: this was probably due to the weakness of the emission and insufficient instrumental sensitivity. Here, with improved sensitivity, a clear difference was observed in the spectral shape between the emission in an alkaline solution and that in an acidic one (Fig.1). It can be seen that the spectrum in an acidic solution is very broad and shifted to the red com- pared with that in an alkaline solution; the emission yield in acidic solution is reduced by about an order of magnitude. For the 4-PPK, the spectral shape of the emission was maintained within the pH range where it could be detected (PH = 4-10). Transients detected by Laser Flash Photolysis The transient absorptions, observed immediately after laser excitation of alkaline aqueous solutions of 3-and 4-PPKs, were securely assigned to triplets, based on comparison with literature values for these moleculesg and benzophenone'0-'2 and by matching of the lifetime values with those obtained by phosphorescence measurements. Also, the rate constants for quenching by 0, [k,, = (1-2) x lo9 dm3 mol-' s-'1 were of the order of magnitude expected for carbonyl tri~1ets.I~ As the acidity increased, spectral changes were accompa- nied by a decrease in transient lifetimes.No notable change in spectral distribution was observed between de-aerated and air-equilibrated solutions over the whole pH range explored. The spectral and kinetic characteristics of 3-and 4-PPK are reported in Table 2. 3-PPK The transient absorption spectra of 3-PPK at three typical pH values in de-aerated solutions are shown in Fig. 2. As can be seen, an intense absorption was always present in the 320-330 nm region, while the spectrum changed at longer wave- lengths. In the visible region, a maximum was observed at 530 nm in neutral and basic solutions.At pH = 2.5-3, the absorption in the visible consisted of two bands of compar-able intensity (A,,, = 505 and 760 nm). With increasing acidity, the spectrum evolved to a maximum at 480 nm, in 1' I I I I I I II 0.0 400 500 600 700 A/nm Fig. 1 Normalised phosphorescence spectra of 3-PPK in aqueous solution at room temperature. 1, pH = 8; 2, pH = 2.7. J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 28 1 Table 2 Spectral and kinetic characteristics of the transients obtained in aqueous solutions of 3-PPK and 4-PPK excited with 347 nm laser excitation 3-PPK 4-PPK 2.7 0.5PH 7.4 11 6.9 5.5 4.1 4naJnm 530 505; 760 480 510 520 515 dps 6.45 2.6; 2.1 1.25 k,,/109 dm3 mol-ls-' 1.9 0.8 2.1 0.29 4.9; 4.7 0.013k: lo9 dm3 mol-' s-l 4.35 1.6 0.36 3.3 0.5 1.8 0.36 0.34 0.25 0.38 a Kinetic rate constants for the quenching by the inorganic anion Cr(CN):- addition to the 325 nm transition.The kinetics of transient disappearance, evaluated at several wavelengths, are shown in Fig. 3. The triplet lifetime (first-order kinetics), measured in basic and neutral solutions (z = 6.45 ps, under the experimen- tal conditions used, c = 1.2 x mol dm-3) decreased as 0.10 the acidity increased, starting from pH 5. In the pH range 3.5-2.5 the lifetime remained constant (t= 2.6 p),then it decreased further to 1.2 ps at pH 0.5. In addition to the short-lived triplet, a long-lived transient, absorbing in the UV region was revealed by time-resolved spectroscopy; this appeared more evident in acidic solutions [Fig.2(6) and (c)]. The absorption maxima (320-330, 360- 0.00 370 and 480-490 nm) and their relative intensity changed a little as the pH changed. Because of the weakness of the absorption signals and partial overlapping with the triplet spectra, the kinetic and spectral analyses were not easily feas- ible. However, in an acidic Ar-purged solution of 3-PPK, the 0.04 AA 10 pst 1i:' .. I 360 nm signal remained constant over a 2 ms timescale, while 0.05 -OCQ -0.00 -!I I , I , I , , I-o.lo-~o' I " ' I ' ' I-kk{ 0.15 0.10 &lo 0.02 0.00 I 0.08 0.04 I I I n, . 1ow .I \' I IIk AA ...-:.. 'W 0.05 ..:.. .. .. ...,.'.?..-..::..:..-..-.._. * .....I. 1 I .::. time 0.00 0.10 0.05 lo% 300 400 500 600 700 800 A/n m Fig. 2 Transient absorption spectra of 3-PPK in aqueous solution. (a) pH = 7.4; (0)0.8 and (A) 9 ps after the flash; (b)pH = 2.7; (0) 0.2 and (A) 3 ps after the flash; (c) pH = 0.5; (0)0.2 and (A) 2 ps after the flash. Fig. 3 Decay profiles of the transients obtained from 3-PPK in aqueous solution. (a) pH = 7.4; 1, 325; 2, 390 and 3, 530 nm; (6) pH = 2.7; 1, 360; 2, 510 and 3, 700 nm; (c) pH = 0.5; 1, 325; 2, 360 and 3,480 nm. in the presence of air it decayed with a first-order rate con- stant k = 1.3 x lo3 s-'. Since no accumulation of these transient(s) could be observed during the triplet decay, we believe that the triplet is not a prescursor. Noting that inter- system crossing yields are less than unity (Table l), these transient(s) could possibly come from the singlet. Photochem- istry was confirmed as a minor decay root by analysis of the So +S, absorption spectra obtained after many shots.4-PPK The absorption spectra of the transients obtained on flashing the 4-PPK at three pH values are shown in Fig. 4 and the 282 J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 0.15 I 0 0.1 0 0.05 00 0.08 p' 0.00 0.10 II I I I I I I I I II I I I- 0.00 I .., . .. .. ... . I I 0 I I AA 0.05 0 0.08 AA 0.00 0.04 0.10 .. _.0 I.. , . .... . ..*:... ...-.... 00 0.05 0.00 300 400 500 600 700 A/n m Fig. 4 Transient absorption spectra of CPPK in aqueous solution.(a) pH = 10; (0)0.8 and (A) 6 ps after the flash; (b)pH = 7.4; (0) 0.3 and (A) 4 ps after the flash; (c) pH = 4.1; (0)0.02 and (A) 0.2 ps after the flash. corresponding kinetic profiles are illustrated in Fig. 5. At pH = 10, the absorption spectrum is very similar to that found in an inert solvent, perfluoromethyl-cyclohexane,2with a high-intensity band at 325 nm and a lower-intensity band centred at 510 nm which consisted of two peaks (505 and 520 nm). The lifetime was 4.35 ps (c = 7 x mol dm-j). At pH = 7 these two peaks coalesced into one at 520 nm and the lifetime decreased to 3.3 ps. At pH = 4, the spectrum showed only a slight change (A,,, = 330 and 515 nm), but the lifetime decreased to 0.25 ps. When the pH was less than 4, the triplet absorption was below our detection limits, due to further decrease in lifetime.Even in this case, photochemistry was observed as a minor process. Excited-state Acid-Base Titrations In order to obtain information about the excited-state basi- city of these molecules, phosphorescence intensity, triplet life- time and triplet-triplet absorption were measured as a function of pH in the range 0-10. The results of the excited- state titrations, from both phosphorescence intensity and T-T absorption (delay time: 50 ns), are compared with ground-state spectrophotometric titrations in Fig. 6 and 7 for 3-and 4-PPKs, respectively. In both titration plots of excited-state 3-PPK, two inflec- tion points were observed, while, for excited-state 4-PPK, the decrease in phosphorescence intensity and T-T absorption was spread over a large pH range.A shift to lower pH values of the T-T absorption titration curve, with respect to the phosphorescence titration curve, was observed for both mol- ecules. For 3-PPK, two inflection points are present, while, 0.00 I I I I I I 0.08 0.04 0.00 time Fig. 5 Decay profiles of the transients obtained from CPPK in aqueous solution (a) pH = 10; 1, 325; 2, 350 and 3, 515 nm; (b) pH = 7.4; 1, 330; 2, 350 and 3, 510 nm; (c) pH = 4.1; 1, 330; 2, 360 and 3,510 nm. for 4-PPK, T-T absorption and phosphorescence emission decrease continuously over a large pH range. For both mol- ecules, the difference between excited-state and ground-state titration indicates a change in basicity upon excitation.-8 -4 0 4 8 Ho PH -+--+ Fig. 6 Phosphorescence (0)and T, +T, (A) titration curves of 3-PPK compared with ground-state spectrophotometric titrations (0).(Ho: acidity function, from M. A. Paul and F. A. Long, Chem. Reu., 1957,57, 1.) J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 I‘ 7 h .-v)4-5 1.0 I I -a -4 0 4 a Ho + PH + Fig. 7 Phosphorescence (0)and T, -+ T, (A) titration curves of CPPK compared with ground-state spectrophotometric titrations (0).(H,,: acidity function as Fig. 6.) Triplet lifetimes, measured as a function of pH, followed the same trend as phosphorescence titrations. Discussion Excited-state pK*s The ground-state dissociation constants of the pyridinium cations derived from PPKs were previously determined (pK3-PPK= 3.0 and pK,_,,, = 3.1).4 Those for the carbonyl protonated molecules (pK < -6) were smaller than that for benzophenone (pK = -4.74;12 -6.1;13 -5.714) because of the electrostatic repulsion by the protonated nitrogen.Relative phosphorescence intensity us. pH plots are expected to yield sigmoid curves. The inflection point pH cor- responds to pK* only when the equilibrium is reached and the lifetimes of the conjugate acid-base pair are identi-cal 15.16 pK* = pH + log zacid/zbase (1) The relative emission intensities, Zo/Z, should fall with increasing acidity according to the equation :‘ = 1 + k,Tb[H+]/(1 + kzt,) (2) where k, and k, refer to the bimolecular interaction of the basic species with the proton and dissociation of the proto- nated species, respectively, and zb and z, are the correspond- ing lifetimes.Linearity of the Zo/Z us. pH plots was observed only within limited pH ranges: 7-5.5, for 3-PPK, and 8.5-7, for 4-PPK. The slopes, divided by zb (the lifetime measured in alkaline solution), gave kJ(1 + k,z,) values (3 x lo9 dm3 mol-’ for 3-PPK5 and 3 x 10l2 dm3 mol-’ for 4-PPK) which were larger than that for benzophenone (4 x lo8 dm3 mo1-I). Owing to the relatively long lifetime of the triplet state, which generally allows acid-base equilibrium in the excited state to be attained, the inflection point in the triplet-triplet absorption titration should give the pK* directly.” Since the molecules absorb at the wavelength dictated by the laser, the relative amounts of acid and base initially excited, in the pH regions where the ground-state equilibria are established, are different from the ground-state acid :base ratio, because of the difference in their absorption coefficients.However, mea- surements with 308 nm laser excitation = 2.1, for 3-PPK, and 2.3, for 4-PPK) and 347 nm laser excitation 283 (Eacid/&base = 1.1, for 3-PPK, and 1.5, for 4-PPK) gave similar results. This, generally, denotes equilibrium attainment during the excited-state lifetime. If this is the case, the inflec- tion points in the T, -+ T, titrations (for 3-PPK, at pH = 3.7 and 1 .O; for 4-PPK, at pH = 4.1) could be taken as a measure of the pK*s.The occurrence of the inflection points at higher pH values in the phosphorescence titration curves than in the triplet- triplet absorption titrations indicates that the protonated forms are shorter-lived species than the basic ones, in the excited state.” From the shift between the phosphorescence titration curve and the T, -,T, titration curve, the lifetime ratio (z,,id/zbase) can be obtained on the basis of eqn. (1). This ratio is less than 0.02 for the 3-PPK (first inflection point, pK* = 3.7 & 0.2 from the laser measurements) and less than 0.001 for the 4-PPK. For the second protonation step of 3-PPK (pK* = 1.0 f0.5, from the T-T absorption titration), the inflection point was determined with less precision in both phosphorescence and triplet-triplet titrations, because of the weakness of the signals and, therefore, the small difference between the two titrations (0.5 pH units) can only be con-sidered as indicative of similar lifetime of the equilibrating species.Lifetime data did not fit the Ware’* and/or Wyatt” equa- tions which would be an alternative way to determine the excited state pK*. This may indicate either the occurrence of multiple acid-base equilibria or non-attainment of the equi- librium in the excited state. Transient Assignments To assign the absorptions, observed under laser excitation, to the triplet species which were involved in the acid-base equi- libria, the principal information was derived from the com- parison of phosphorescence emission with laser data and from charged-ion quenching experiments.The first point is that room-temperature phosphorescence will come exclusively from triplet n,n* excitation localized on the carbonyl. When the carbonyl protonation occurs, the n,x* emission is expected to disappear, as is also observed for benz~phenone.’.’~ In addition, two emitting species, the neutral molecule and the pyridinium cation, are expected because of the presence of the nitrogen atom. Thus, for the 3-PPK, the two phosphorescence emissions, shown in Fig. 1, are assigned to the carbonyl triplet (n,n*) of the neutral molecule (in alkaline solution) and to the triplet (n,n*) of the pyridinium cation (recorded at pH = 3).Analo-gously, the T, +Tn absorption spectra, detected in these pH regions [spectra (a) and (b) in Fig. 21, are assigned to the neutral triplet and to the nitrogen-protonated triplet, respec- tively. These species are characterized by triplet lifetimes in the ratio z,/q, x 0.4, which greatly exceeds that found from the relative shifts of the emission and absorption titration curves (z,/zb < 0.02). Consequently, the excited-state pK* determined from the titration curves does not refer to equi- librium establishment between the triplet molecule and the triplet pyridinium cation. The triplet molecules equilibrate with a shorter-lived species (not revealed by absorption or emission spectroscopy), which we assign to the carbonyl protonated cation.When the pyridinium cation is directly excited (ground-state pK = 3), the positive charge on the nitrogen atom opposes carbonyl protonation. Thus the absorption and emission signals from this species are con- stant within a limited pH range (2-3). Further decrease of pH induces carbonyl protonation and the last absorption spec- trum in Fig. 2 (pH = 0.5) corresponds to the dication, which does not show room-temperature phosphorescence emission. 284 In order to ascertain the cationic nature of the transients observed in acidic solution and to decide whether the two bands observed around pH x 3 for the 3-PPK belong to the same species, quenching experiments were performed using a negatively charged ion, Cr(CN);-, as a quencher.The triplet energy of this anion (ET= 160 kJ mol-') is lower than that of triplet ketones (see Table 1) so as to ensure diffusional energy transfer. Based on the Debye equation [eqn. (3)],19 k:irf = (BNk, T/3000q)a[exp(a) -13-' (3) a = ZDZQe2/r&kB (34T when both the donor (D) and quencher (Q) are ionic species, the diffusional constant depends on the product of their charge, ZDZ,. N and k, are the Avogadro's and Boltzmann's constants, e is the electron charge, q and E are the viscosity and relative permittivity of the medium and r is the encoun- ter distance. The influence of the ionic strength, p, on the rate constant can be evaluated by the Debye-Brmsted equation (4) log kdiff = log k,Oiff + 1*02Z,z, JPI(1 + JP) (4) Quenching measurements of the triplet lifetime of 3-PPK (D) by Cr(CN):- (Q) were carried out at the three pH values which corresponded to the neutral, mono- and di-cationic species, on the basis of the above assignment.Quencher con- centrations were adjusted to have at least 50% of quenching at each pH value. Linear Stern-Volmer plots of T-' us. [Q] were obtained. From the data reported in Table 2, it can be seen that, for the 3-PPK, the quenching rate constant increases on going from neutral solution (k = 2.9 x lo8 dm3 mol-' s-') to pH = 2.7 (k = 4.8 x lo9 dm' mol-' s-') and then decreases again when pH 0 is approached (k, = 1.3 x lo7dm3 mol- ' s-'). It is worthwhile noting that the same rate parameter was found for the 505 and 760 nm bands at pH = 2.7.These results are consistent with the prevalent presence, depending on the acidity of the solution, of a neutral species (pH = 7), a mono-cationic species (pH x 3) and a di-cationic species (pH x 0.5). The quenching rate parameter of the neutral triplet transient is somewhat lower than expected from a diffusion-controlled process (6.6 x lo9 dm3 mol-' s-') but it is of the same order of magnitude as those found with other organic molecules as donors.20 Non- diffusional k, values were attributed to shielding of the ionic quencher by polar solvent molecules.20 The value measured at pH = 3 is higher because of the attractive electrostatic force between PPKH' and the inorganic anion and is in good agreement with the calculation from eqn.(3) and (4) (kdiff = 5.9 x 109 dm3 mol-' s-', at the experimental p = 0.011), when an encounter distance of 15 A is assumed. Even though the electrostatic interaction increases with the di-cation, the rate constant decreases because of the ionic strength effect in the strongly acidic medium (p= 0.79). The calculation for ZDZQ = -6 leads to kdiff = 2.6 x i07 dm3 mol-' s-'. The agreement with the experimental value is satisfactory, considering the approximations inherent in the calculation and the large difference with the calculated value of the mono-cation interaction in this medium (ZDZQ= -3, kdiff = 4.5 x 108 dm3 mol-' s-I). For 4-PPK, the phosphorescence measurements indicate the presence of a unique emitting species, which we identify as the triplet of the neutral molecule.Quenching experiments, carried out at four pH values with Cr(CN);-as a quencher, did not show any particular evi- dence of the presence of cationic species for 4-PPK. The quenching rate parameter measured [k, = (3.6 f0.2) x lo* dm3 mol-' s-'1, which is very close to that found for the neutral triplet of the 3-PPK, is constant, at constant ionic J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 strength (p= 0.011), in the pH range 4-10. Thus, it can be concluded that it always refers to the unprotonated triplet molecule. This does not mean that protonation does not occur, but, owing to the very short lifetime of the protonated form(s), only the neutral species is quenched, even if it is present as a very small fraction at the lowest pHs explored.Therefore, the T, + T, absorption spectra, which could be observed within the same pH limits, also belong to the uncharged triplet species. The negligible differences observed in the spectral distribution (Fig. 4) are within the measure- ment uncertainty. Triplet lifetime and absorption intensity are reduced because of interaction with the proton as the acidity increases. If we recall that the lifetime difference between neutral and protonated forms, evaluated from the shift between the phosphorescence and triplet-triplet absorp- tion titration curves, is larger than three orders of magnitude, T,,id should be less than 0.004 ps Thus, the triplet of the protonated form cannot be revealed either from quenching measurements or from its absorption spectrum with our experimental set-up.The evidence of only one distinct emis- sion spectrum, which became undetectable below pH = 4, not only denotes that protonation occurs first on the car- bony1 in the excited state, but also that it is immediately fol- lowed by nitrogen protonation, since no pH value could be found where the pyridinium mono-cation could be detected. Double protonation should be easier for 4-PPK than for 3-PPK, due to the larger distance between the two proto- nation sites. Thus, owing to the closeness of the two pK* values, it was impossible to separate the two protonation steps. Therefore, the inflection point in the titration curve, which is due to two overlapping protonation steps, probably does not correspond to a real pK*. The comparison with benzophenone on the basis of eqn.(2) points to a marked difference in lifetime of the acidic forms (TB > T3-ppK % Tq-ppK). Concluding Remarks The results of this work demonstrate further that the excited triplet carbonyl is a much more basic species than the ground state. For both these molecules, the nitrogen effect is that of enhancing the basicity of the triplet carbonyl compared with benzophenone. This result was unexpected since an electron- withdrawing group, such as pyridine, should decrease the pK*. However, an analogous behaviour was found for the dissociation constants of the ketyl radicals derived from PPKs with respect to the diphenylketyl radical.21 The increased basicity of benzophenone upon excitation was attributed to the intramolecular charge-transfer character of the T, state in polar media.22 In the present cases, charge transfer from the pyridyl ring to the carbonyl should be more efficient than that from the phenyl ring as previously pro- posed for the di-pyridyl ketone^.^ The stabilization of n,n* states on introducing the nitrogen atom in pyridyl ketones compared with benzophenone, that can be deduced from the bathochromic shift of the n,n* transitions, supports this inter- pretation.These two molecules exhibit very similar photophysical properties in organic solvents as well as in alkaline aqueous solutions. In contrast, they are well differentiated in acidic media where the protonated species could be spectrally dis- criminated for the 3-isomer only.The very short lifetime pre- vents any spectral evidence of the 4-PPK acidic form from being obtained by either emission and absorption spectros- copy (nanosecond time resolution) and, probably, makes it impossible to obtain acid-base quilibrium in the excited state. We thank Prof. Riccieri for the sample of K,Cr(CN),. The financial support of the Italian Consiglio Nazionale delle J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 Ricerche and Minister0 per l'Universita e la Ricerca Scienti- fica is gratefully acknowledged. References 1 P. Traynard and J. P. Blanchi, Mol. Photochem., 1972,4,223. 2 J. P.Blanchi and A. R. Watkins, Mol. Photochem., 1974,6, 133. 3 C. R. Hurt and N. Filipescu, J. Am. Chem. SOC., 1972,94,3649. 4 G. Favaro, J. Chem. SOC., Perkin Trans. 2, 1976, 869. 5 G. Favaro and F. Masetti, J. Phys. Chem., 1978,82, 1213. 6 G. Favaro, F. Masetti and A. Romani, Spectrochim. Acta, Part A, 1989,45339. 7 A. Romani, F. Elisei, F. Masetti and G. Favaro, J. Chem. Soc., Faraday Trans., 1992,88,2147. 8 G. Favaro, J. Photochem., 1986,33,261. 9 S. Monti, N. Camaioni and P. Bortolus, Photochem. Phorobiol., 1991,54,325. 10 M. B. Ledger and G. Porter, J. Chem. Soc., Faraday Trans. I, 1972,683,539. 11 R. V. Bensasson and J. C. Gramain, J. Chem. Soc., Faraday Trans. I, 1980,77, 1801. 285 12 D. M. Rayner and P. A. H. Wyatt, J. Chem. Soc., Faraday Trans. I, 1974,70,945. 13 A. Fischer, B. A. Grigor, J. Packer and J. Vaughan, J. Am. Chem. Soc., 1961, 83, 4208. 14 T. G. Bonner and J. Phillips, J. Chem. SOC. B, 1966,650. 15 J. F. Ireland and P. A. H. Wyatt, Ado. Phys. Org. Chem., 1976, 12, 1931. 16 A. Samanta, N. Chattopadhyay, D. Nath, T. Kundu and M. Chowdhry, Chem. Phys. Ltt.,-1985,121, 507. 17 G. Favaro and G. Bufalini, J. Phys. Chem., 1975,80,800. 18 W. R. Ware, D. Watt and J. D. Holmes, J. Am. Chem. Soc., 1974, 96,7853. 19 P. Debye, Trans. Electrochem. Soc., 1942,82,265. 20 H. F. Wasgestian and G. S. Hammond, Theor. Chim. Acta, 1971, 20, 186. 21 D. A. Nelson and E. Hayon, J. Phys. Chem., 1972,76,3200. 22 H. Shizuka and E. Kimura, Can. J. Chem., 1984,62,2041. 23 H. Shizuka and E. Kimura, Can. J. Chem., 1984,62,2041. Paper 3/03114D;Received 1st June, 1993