1924
J . Phys. Chem. 1986, 90, 1924-1927
be in the range from 8.8 X 1O'O to 2.66 X 10'' M-' s-', depending on the viscosity and dielectric constant of the solvent." It has been shown that the normalization of the rate constant, k2', for viscosity, q, and dielectric constant, e , results in the same value; kieq = 7.4-8.1. The kZeqvalues for the reactions of the radical anions are in the range from 9.1 to 14 when we take q = 0.46 and c = 7.58 for THF. It is then said that the normalized rate constants for the neutralization reactions of the radical anions with THF(H+) are larger than that of benzhydryl cations with C1-. This seems reasonable if we consider that these neutralization reactions are diffusion controlled and that THF(H+) is similar to H+ in its diffusion behaviors. The magnitude of the rate constants also demonstrates that the mobility of the reacting species is increased by the attractive force of the coulombic interaction. The rate constants for the Bu4NPF6 solutions are one to two orders of magnitude smaller than those for the salt-free solutions. (17) Kundu, K. P.;Dorfman, L. M. Radiat. Phys. Chem. 1982.20, 247.
In the presence of the salt, the neutralization reactions are considered to occur between the ion pairs (radical anions)/Bu4N+ and THF(H+)/PF6-. The decrease in the rate constant by ion pairing is attributable to decreases in the mobility and reactivity of the reacting species. The effect of the coulombic interaction is not important for the ion pairs. The difference in the rate constants among the radical anions is large for the Bu4NPF, solutions compared with that for the salt-free solutions. Furthermore, it is too large to be explained in terms of the difference in mobility among the radical anions. Therefore, it is considered that the neutralization between the ion pairs is not necessarily a diffusion-controlled reaction. The rate constant for the ion pairs decreases in the order BP-. > St-. > An-. > Py-., and the degree of charge delocalization of the radical anions may be responsible for the difference in the rate constants.
Acknowledgment. We are grateful to Mr. Kunihiko Tsurnori, Mr. Norio Kimura, Mr. Tamotsu Yamamoto, Mr. Toshihiko Hori, and Dr. Seishi Takeda for help with the pulse radiolysis experiments.
Excimer Formatlon from Triplet-Triplet Annihilation Reactions of the Lowest-Lying Triplet Excited State in Aluminum( I I I ) , Silicon( I V ) , and Metal-Free Phthalocyanines: Medium and Magnetic Field Effects on the Rate of Reaction M. E. Frink, D. K. Geiger, and G. J. Ferraudi* Radiation Laboratory, University of Notre Dame, Notre Dame, Indiana 46556 (Received: August 26, 1985)
The relaxation kinetics of aluminum(III), silicon(IV), and metal-free phthalocyanine (PcH,) have been investigated by both conventional and laser flash photolysis. The observed excited-state decay kinetics are shown to be controlled by the excited-state concentration, the nature of the solvent, and the presence of an external magnetic field. Triplet-triplet annihilation is shown to be a dominating factor in the excited-state scheme. The second-order deviations observed in the relaxation kinetics of the phthalocyanines under various conditions are discussed in terms of an excimer- ((3E,)Pc* + (3E,)Pc* ('A,,)[Pc-Pc]*l mediated mechanism.
-
Introduction
A considerable amount of work has been reported on the
properties of the lowest-lying triplet of the metallophthalocyanines.'-* Such species exhibit lifetimes that are dependent on the metal center and vary from a few nanoseconds, e.g. for copper( 11) phthalocyanines,6 to several hundred microseconds, e.g. for aluminum(II1) phthalocyanines.'** In the redox quenching of the excited state, electron transfer proceeds by an outer-sphere mechanism with the excited phthalocyanine working as an acceptor or a donor of electron^.^,^ Furthermore, some quenching reactions are mediated by exciplexes formed when excited phthalocyanines react with ground-state aromatic q u e n ~ h e r s . ~ Ohno et aL5 have recently reported that a triplet-triplet annihilation reaction contributes to the rate of 3E, decay in alu(1) Darwent, J . R.; McCubbin, I.; Phillips, D. J. Chem. Soc., Faraday
Trans. 2 1982, 78, 341. (2) Ferraudi, G . ; Muralidharan, S. Inorg. Chem. 1983, 22, 1369. (3) Prasad, D. R.; Ferraudi. G . J . Phvs. Chem. 1982. 86. 4037. (4) Ferraudi, G.;Prasad, D. R. J. Chem. SOC.,Dalton Trans. 1984, 2137. ( 5 ) Ohno, T.; Kato, S.:Yamada, A.; Tanno, T. J. Phys. Chem. 1983, 87, 77s. ( 6 ) Prasad, D. R.: Ferraudi, G. Inorg. Chem. 1982, 21, 2967. (7) The phthalocyanines can be considered to belong to D4hor D, point groups, an approximation that has been already used for the evaluation of electronic energiess This is not totally correct with metal-free phthalocyanine where the protonation of the ligand removes the fourfold symmetry axis. For a D,, symmetry, the lowest-lying excited states Can be assigned as 'E, and 3E,. The u denomination will be lost in a D4point group, (8) Schaffer, A. M.; Gouterman, M.; Davidson, E. R. Theor. Chim. Acta 1973, 30, 9. (9) Prasad, D. R.; Ferraudi, G Inorg. Chem. 1983, 22, 1672.
0022-3654/86/2090-1924.$01.50/0
minurn(II1) phthalocyanine. Although these authors found no spectral evidence for the formation of redox products in the annihilation process, i.e. phthalocyanine radicals, a more recent study has revealed that excimers participate as reaction intermediates.l0 Despite these studies little is known about the mechanism of triplet-triplet annihilation and the significance of such a process on the photophysics of excited phthalocyanines. We have investigated the possible formation of radicals in the decay of phthalocyanine excirners in protic and aprotic solvents. Moreover, the reactivity of the excited species in phthalocyanines which possess long-lived triplet states has been studied by affecting the reaction rate with intense magnetic fields. Experimental Section
Photochemical Procedures. Photochemical reactions were investigated on a microsecond-millisecond time domain with a flash photolysis apparatus described elsewhere.* In these experiments, the sample was irradiated with a pulse of light generated by firing two Xe flash lamps at appropriate energies. The light was filtered with a band-pass filter in order to irradiate only at wavelengths of the phthalocyanine Q band, e.g. 600-700 nm. For the study of reaction kinetics or the acquiring of spectral information, the phototube response to changes in the intensity of the probing beam was digitized in a transient recorder, Biomation 805, and processed in a PDP 11/55 computer. The effect of the magnetic field on the rate of reaction was investigated by laser flash photolysis in an adaptation of a pro(10) Frink, Mark E.; Ferraudi, G . Chem. Phys. Leu., in press
0 1986 American Chemical Society
nloo
The Journal of Physical Chemistry, Vol. 90, No. 9, 1986 1925
Phthalocyanine Relaxation Kinetics
c .
I.""
0.7
0.6. 0.5.
1.08
0.4 .
t
-
0*3
I
U
I
I
I
I
I
0.141
0.12 0.10
380.0
-
456.7 495.0
533.3
571.7
610.0
WAVELENGTH
Figure 2. Transient spectra recorded in flash irradiations A(,
-
of
0.08 0.06
418.3
2 540 nm)
M Si(Pc)(OEt), in deaerated cyclohexane.
TABLE I: Typical Dependence of the Excited-State Half-Life on Excited-State Concentration
-
compd Si(PC)(OEt),
t
0*04
406
500
600
WAVELENGTH, nm
Figure 1. Transient spectra recorded in flash irradiations (A,
2 540 nm)
of 5 X lod M AI(Pc)CI in deaerated ethanol. The spectra were obtained with two different delays, t = 50 ps and r = 800 ps, in order to show the transformation of the excited-statespectrum ( t = 50 ps) into the spectra of phthalocyanine radicals ( t = 800 ps). cedure previously described.I2 Appropriate dyes were used with a flash-lamp-pumped dye laser, Candela SLL-66, in order to produce pulses of light of around 640 nm. The sample, placed in the cavity of a solenoid, was immersed in a magnetic field that was generated by discharging a capacitor across the coil. Fields whose peak intensities remained nearly constant for ca. 3 ms and were adjustable between 0 and 70 kG could be obtained with this experimental setup. In each experiment, the laser pulse and the magnetic field were recorded on an oscilloscope in order to verify the synchronization between the magnetic field and the laser pulse as well as the constancy of their corresponding intensities. All the flash photolysis work was carried out with solutions deaerated with streams of Ar. Materials. The nearly pure phthalocyanines available from Eastman, Al(pc)Cl and pcH,, were recrystallized five times from ethanol and dried under vacuum.*I The silicon dichlorophthalocyanine, Si(pc)C12, available from ICN Pharmaceuticals was converted to the diethoxy complex, Si(pc)(OEt),, by reported procedures.13
Results The kinetics of relaxation of the excited phthalocyanines have been investigated as a function of the excited-state concentration, the nature of the solvent, and the presence of magnetic fields. The corresponding observations are reported below. Medium and Concentration Effects. Transient spectra were determined from flash photolysis (A,, = 650 nm) of Al(pc)Cl, Figure 1. Irradiations of deaerated solutions of the complex in methanol or acetonitrile gave the same transient spectra at short times ( t < 100 ps); these spectra are in agreement with the one (1 1) Abbreviations: The silicon diethoxy- and dichlorophthalocyanines, Si(pc)X, with X = OEt or C1, the aluminum chlorophthalocyanine, Al(pc)Cl, and the metal-free phthalocyanine, pcH2, are generically described as Pc in the chemical equations. (12) Pacheco, M.; Ferraudi, G . Chem. Phys. Lett. 1984, 222, 187. (13) Krueger, P. C.; Kenney, M. E. J . Org. Chem. 1963, 28, 3379.
Al(Pc)Cl PcH~
tl12, C ~ S
~
0
81 115 124 132 139 159 187 113 132 156 63
1.45 1.05 0.80 0.60 0.48 0.30
87
1.01 0.21
156 74 92 117
~
solventa 4 ethanol
0.20
1.40 0.80
cyclohexane
0.20
1.37
1S O
ethanol cyclohexane
0.80 0.20
"Change of the optical density at, , ,A after the flash irradiation 2 570 nm) of solutions deaerated with streams of 02-free nitrogen. The AOD was determined in cells with a 20-cm optical path.
A(,,
reported for the lowest-lying 3E,. Also, the transient spectra recorded from flash photolysis of Si(pc)(OEt), in either ethanol or cyclohexane and pcH2 in chloronaphthalene correspond to the spectrum of the jE,. With all these compounds, the transient spectrum of the excited state evolves with time into the combined spectra of the phthalocyanine cation and anion radicals, e.g. Al(pc)C1'- and Al(pc)C1'+, when large concentrations of the excited state are generated in flash irradiations of the parent phthalocyanines in protic solvents, Figure 1. However, no such radicals are detected in aprotic solvents such as cyclohexane, Figure 2. We have verified that the disappearance of the excited state in these phthalocyanines depends on the solvent and on the initial concentration of the excited state, Table I. For large concentrations of the excited state in protic solvents, the rate of disappearance departs from first-order kinetics, but the rate approaches it at low concentrations, Figure 3. Althaugh in aprotic solvents the decay of the excited phthalocyanines does not form radicals, the rate law deviates from first order when large concentrations of excited state are produced in flash irradiation^.'^ The dependence of the reaction rates on the initial excited-state concentration and curve-fitting indicate that the decay of the excited state can be regarded as a competition between first- and second-order reactions. Rate constants for the two decay reactions (14) Small spectral changes observed during the decay of the excited phthalocyanines in aprotic solvents can be related to the formation of intermediates as reported for various porphyrins."
1926 The Journal of Physical Chemistry, Vol. 90, No. 9, 1986
Frink et al.
d
:t O 040 0
o
6
0
r
W
0
z
3a
8a
I.!
TIME (MICROSECONDS)
1
0.201
-'a -2 U-
I.(
I
c
W
Y 3 8 v)
m 4
0.' TIME (MICROSECONDS)
Figure 3. Traces for the decay of the excited state of Si(Pc)(OEt), under a regime of low excited-state concentration (above) and high excited-state concentration (below). Notice deviations from a first-order fit at a high concentration of excited state and the opposite at a low concentration of N 490 nm. For other excited state. Both traces were obtained at A,, conditions see Figure 2. TABLE 11: First-Order ( k , ) and Second-Order ( k 2 )Rate Constants for the Decay of the Lowest-Lying Excited State in Phthalocyanines
compd AI (Pc) Clb
Si(Pc)(OEt),' PcH,'
k,/103,"s-] 9.0 f 0.2 9.8 f 0.2 3.5 f 0.1 4.8 f 0.1 7.8 f 0.3 9.6 f 0.2
(k2/e)103," cm-' s-I 18.2 f 0.5 13.0 f 0.5 53.5 f 0.8 20.0 f 0.8 1 5 . 3 f 0.6 40.2 f 0.6
H , kG 0 70 0 70 0 70
Figure 4. Dependence of the excited-state half-life T , on the initial concentration of excited state in the absence and presence of a 70-kG magnetic field. The laser flash experiments with PcH, and Si(Pc)(OEt), were carried out in deaerated 1-chloronaphthalene and ethanol, respectively. The equation in the insert is the expression of the reaction half-life for less than a few reactions percent (initial rate approximation). The rate constants k l and k2 correspond to the overall rate constant for the radiative plus nonradiative relaxation of the excited state and the excimer formation respectively.
"The values of k , and k , correspond to the intercept and slope, respectively, from a least-squares fit of the experimental data plotted as in Figure 4 The decay of the excited phthalocyanines was followed at Amx. * 5 X 2c 6 X 10" M complex in deaerated CH,CN. ' 5 X M complex in deaerated I-chloronaphthalene. 2 c 2 6 X
-
of the excited states of Al(pc)Cl, Si(pc)(OEt),, and PcH2 are reported in Table 11. Kinetics of Luminescence. That the second-order reaction must be described as triplet-triplet annihilation is in agreement with the observation of a long-lived emission at wavelengths where Si(pc)(OEt), exhibits prompt fluorescence, i.e. at 600-750 nm. In deaerated benzene or cyclohexane, such a delayed emission, studied by flash fluorescence, decays with the same lifetime as the corresponding triplet state, i.e. t , / , 90 ps in emission and absorption experiments. Magnetic Field Perturbation. The effect of an applied magnetic field on the rate of disappearance of the 3E, was investigated by laser flash photolysis as a function of the excited-state concentration. Application of a magnetic field increases the rate of the first-order (radiative and nonradiative) relaxation and reduces the rate of the second-order triplet-triplet annihilation, Figure 4 and Table 11. Moreover the field also increases the yield of triplet state, e.g. nearly a 10% increase for a 70-kG field.
4 E N
E
R 0 Y
-
Discussion The results reported above show that the triplet-triplet annihilation of 3E, phthalocyanines produces phthalocyanine radicals
REACTION COORDINATE Figure 5. Hypothetical representation of the potential curves along the
reaction coordinate for the formation of the excimer from the encounter complex (see equation at the top of the figure) in the absence ( H = 0) and presence ( H # 0) of the magnetic field. For H # 0, 2 B H [ = AE, which represents the perturbation introduced by the magnetic field H.
The Journal of Physical Chemistry, Vol. 90, No. 9, 1986 1927
Phthalocyanine Relaxation Kinetics in protic solvents. The formation of radicals with such a fast reaction rate cannot be described as a disproportionation of the excited state, eq 1, which obeys the Marcus mechanism for (3E,)Pc*
+ (3E,)Pc*
-
(2E,)Pc'-
+ (2A,)Pc'+
(1)
outer-sphere electron-transfer reaction^.^-^ Indeed, the known redox potentials for the oxidation and reduction of phthalocyanines show that the production of radicals must be slightly exoergonic in protic solvents, i.e. by less than 0.03 V, and endoergonic in aprotic solvents.l5I8 For such potentials, the treatment of reaction 1 in the same manner used for other outer-sphere electron-transfer reactions of phthalocyanines gives rate constants, k < lo6 M-I s-l, which are several orders of magnitude smaller than the corresponding experimental values."JF21 In this context, it is possible to attribute the solvent-controlledsecond-order decay of the excited state, i.e. the path responsible for radical generation and delayed fluorescence,-to an excimer-mediated mechanim, eq 2-6.,, ('A,,)Pc
hv
;=t
('E,)Pc*
-
(3E,)Pc*
('A,,)[Pc-Pc]*
(lAIg)[P~-P~]*
+ ('A,,)Pc
('E,)Pc*
F?
(3E,)Pc*
(IA,,)Pc
(3E,)P~*+ (3E,)P~* ('A~,)[Pc-Pc]*
-
(*E,)Pc'-
+ (2Al,)Pc'f
(6)
Such a mechanism is consistent with the tendency of the lowest triplet state in phthalocyanines and porphyrins to form exciplexes in quenching reaction^^*^^ as well as with the perturbation of the rate of reaction by the magnetic field.25-30 Indeed, the mag(1 5) The observed change in redox potential from aprotic to protic solvents is in agreement with an increase in the stability of the charged products of the reduction and oxidation reactions (phthalocyanine radicals Pc'- and Pc'*) as the solvation energy of these species is increased. (16) Wheeler, B. L.; Nagasubramanian, G.; Bard, A. J.; Schechtman, L. A,; Dininny, D. R.; Kenney, M. E. J . Am. Chem. SOC.1984, 106, 7404. (17) Lexa, D.; Reix, M. J . Chim. Phys. 1974, 71, 511. (18) Geiger, D.; Ferraudi, G., work in progress. (19) Electron-transfer reactions were treated according to the Marcus20 and Hush2' mechanism for outer-sphere electron-transfer reactions. An upper limit for the rate constant, k 5 lo6 M-' s?, was obtained by using the expression, k12= (kllk22K1fi1/2with logf= (log !12)~/(4 log k,,k22/.Z2),for the cross-electron-transfer reactions. The potentials for the calculation of K and the self-exchange rate constants, k , , and k,,, have been reported elsehere.^,^
(20) Marcus, R. A. Discuss. Faraday SOC.1960, 29, 21. (21) Hush, N. S. Trans. Faraday SOC.1961, 57, 557. (22) The mechanism of triplet-triplet annihilation proposed in this work is in good agreement with the previous report of delayed fluorescence in metal-free phthal~yanine.~'Moreover, the back reaction of the phthalocyanine radicals, eq 6, is similar to one proposed by Bard et al. in connection with the observation of electroluminescence.I6 (23) Stadelman, H. R. J. Lumin. 1972, 5, 171. (24) Whitten, D. G.; Roy, J. K.; Carroll, F. A. The Exciplex, Gordon, M., Ware, W. R., Eds.; Academic Press: New York, 1975; pp 247-271. (25) The variation of the rate constant of a reaction in a magnetic field can be treated in the same manner that Freed and Jortner26used for radiationless transitions and Dexter2' for energy-transfer reactions. In this approach, the Zeeman Hamiltonian HI = p R ( k i
+ 23)
can be regarded as a part of the cou ling operator that has to be used in conjunction with Fermi's golden rule2!
in order to introduce (in the expression of the reaction rate constant) a term that contains the dependence on the magnetic field intensity. Several mechanisms for the perturbation of the radiationless relaxation rates by magnetic fields have been already discussed in the l i t e r a t ~ r e . ~ ~ . ' ~ (26) Freed, F. K.; Jortner, J. J. Chem. Phys. 1970, 53, 6272. (27) Dexter, D. L. J . Chem. Phys. 1953, 21, 836. (28) Ballhausen, C. J. Molecular Electronic Structures of Transition Metal Complexes; McGraw-Hill: New York, 1979; pp 67-89, 149-153, 185-189. (29) Atkins, P. W.; Stannard, P. R. Chem. Phys. Lett. 1977, 47, 1 1 3 . (30) Matsuzaki, A.; Nagakura, S. J . Lumin. 1979, 18/19, 115.
netic-field-induced increase of the rate constant for the first-order decay, i.e. radiative and nonradiative relaxation, is in agreement with a perturbation that allows one to bring singlet character to the triplet state. It is possible to show with symmetry arguments that the lowest-lying singlet and triplet E, states can be mixed by the magnetic field; a mixing that is also in agreement with the observed increase in the 3E, yield. Moreover, the decrease in the rate of the second-order reaction can be related to the combined displacement of the potential surfaces, Figure 5 and A p p e n d i ~ . ~ ' As a consequence of such displacement, the surfaces that are stabilized by the magnetic field do not correlate with the final state of the excimer product, and molecules reacting along this path must overcome an activation energy. This will be the case, for example, for molecules reacting along A,, in Figure 5b. If the separation between surfaces with the "wrong symmetry" and the surface that correlates with the excimer product is not too large, thermal activation can promote molecules to the upper surface and a statistical distribution between these close levels will control the rate of reaction, e.g. for reaction along the surface Al, in Figure 5b. In either case the rate of the second-order reaction is expected to decrease with increasing magnetic field intensities. If one ignores presented evidence on the participation of excimers in the triplet-triplet annihilation, it is possible to analyze the reaction in terms of mechanisms, e.g. dipole-multipole energy transfers, that do not require an interaction between reactants as strong as in the formation of e x c i m e r ~ . * ~However, %~~ the expressions for the rate constants of such mechanisms, when used with symmetry arguments similar to those discussed in connection with paths leading to an excimer, signal that the reaction rates must not be as sensitive to the magnetic perturbation as required by our results.27 Indeed, treatment of the triplet--triplet annihilation according to the mechanism of Atkins and Evans for nondegenerate states34resulted in magnetic-field-induced perturbations of the rate constants one order of magnitude smaller than those measured in our experiments.
Acknowledgment. The research described herein was supported by the Office of Basic Energy Sciences of the Department of Energy. This is Document No. NDRL-2754 from the Notre Dame Radiation Laboratory. Appendix A number of rules have been followed for the construction of the Figure 5.25,33,34The overall procedure is outlined as follows as a series of sequential steps. (1) The excimer is considered to achieve stability by a weak electronic exchange mechanism involving the HOMO orbitals of the excited phthalocyanine^.^^ The resolution of the secular equation gives A,, as the ground state and A,, as lowest-lying excited state of the excimer. Approximate estimates of the electronic energies were obtained from the values reported elsewhere for the monomers.8 (2) The formation of the excimer can be regarded as a fusion process (within the encounter complex) that involves a totally symmetric nuclear displacement as a reaction coordinate. In this context, the C, axes of the reactants are oriented along a common z coordinate. (3) The encounter complex, formed with two 3E, states, can be in either A,,, A2,, B,,, or B,, states which are nearly degenerate in the presence of weak magnetic interactions, Figure 5a. The intense magnetic perturbation removes such a degeneracy, and the treatment of the perturbation by first-order perturbation theory shows that A,, is stabilized by the same energy that destabilizes A l p Figure 5b. The same relationship exists between B,, (stabilized by the field) and Big. (31) For this argument, we have considered an excimer with nearly D4 symmetry which achieves stability through an electronic exchange mechanism, see Appendix. (32) Woodward, R. B.; Hoffmann, R. Angew. Chem., Int. Edit. Engl. 1969, 8, 781. (33) Longuet-Higgins, H . C.; Abrahamson, E. W. J . A m . Chem. SOC. 1965, 87, 2045. (34) Atkins, P. W.; Evans, G. T. Mol. Phys. 1975, 29, 921.