Journal of Luminescence 26 (1982) 297-317 North-Holland Publishing Company 297 RADIATIONLESS INTERMOLECULAR ENERGY TRANSFER WITH PARTICIPATION OF ACCEPTOR EXCITED TRIPLET STATES G.P. GURINOVICH, E.I. ZENKEVICH and E.I. SAGUN Institute of Physics, Academy of Sciences of the Byelorussian SSR, Minsk, B SSR/U SSR Received 17 July 1981 Radiationless energy transfer between like and unlike molecules has been experimentally studied under conditions where acceptor molecules have been excited to the triplet state. Homogeneous singlet-triplet-triplet migration has been discovered in highly concentrated chlorophyll “a” and pheophytin “a” solutions in castor oil at 183 К by measuring the variation of pigment relative quantum yields of fluorescence and triplet state formation as a function of exciting pulse intensity. Heterogeneous singlet-triplet-triplet energy transfer has been observed in soUd solutions of different complex organic molecules (perylene + phenanthrene, Na-fluorescein + chlorophyll “a”, pyrene+ Mg-phthalocyanine) as the fluo­ rescent donor state quenching in the presence of acceptor triplet-excited molecules. Primary emphasis is placed on a direct observation of the effect of energy transfer on the excited-state lifetime of the donor. The benzophenone phosphorescence quenching (shorten­ ing of phosphorescence lifetime) in the presence of Mg-mesoporphyrin triplet molecules has been found to be caused by the heterogeneous triplet-triplet-triplet energy transfer. Good agreement of the theoretical and experimental results permits us to conclude that all types of observed transfer processes are described by the Forster-Galanin theory for dipole-dipole radiationless energy transfer with no additional assumptions. 1. Introduction The majority of experimental investigations of the electronic excitation energy transfer (EEET) in solutions due to the inductive-resonance mechanism of Forster-Galanin [1,2] or electron-exchange interactions of Dexter [3] are devoted to study of electronic energy migration from the lowest electronically excited states of donor molecules (D) to the ground electronic state of acceptor molecules (A) [4,5]. The inductive-resonance theory of EEET is based on the assumption that the leading term in the interaction can be expressed in terms of the dipole-transition moments of D and A, and that the migration efficiency between interacting molecules depends essentially on the allowness of the electronic transitions in A. Provided that the acceptor electronic transition is allowed, the present EEET theories do not exclude the possibility of the 0022-2313/82/0000-0000/$02.75 © 1982 North-Holland 298 G. P. Gurinovich et a i / Radiationless intermolecular energy transfer participation of short living high-excited states of D and A in migration processes. Direct evidence for the radiationless transfer from higher triplet levels of organic molecules to solvent molecules was first obtained by Alfimov and co-workers [6,7]. Later on these authors investigated systematically this phe­ nomenon and analyzed the conditions determining efficiency of such a migra­ tion [8]. Recently the phenomenon of EEET from a highly excited singlet state of the' donor to the singlet ground state of the acceptor was discovered in the crystal matrix [9] and in fluid solutions [10-12]. Kaplan and Jortner [10,11] have extended the Forster-Galanin analysis for the case of EEET from a thermally nonequilibrated vibronic manifold of the donor excited electronic state when the electronic relaxation rate overwhelmed the rate of vibrational relaxation. But in their experiments, the efficiency of the sensitized fluores­ cence of A (2.5 bis (5'-tertbutyl-2-benzoxazolyl) thiophen) as a result of EEET from D molecules (Rhodamine 6G) was very weak due to the low concentra­ tion of A (< 10“ ^M) and the high probability of back transfer. The most direct experimental evidence of energy transfer from highly excited states of D (9.10-bi-n-propylantracene) was obtained by Ermolaev with co-workers [12,13], who discovered the effective sensitized fluorescence of A (naphthalene), caused by using an intermediate acceptor, i.e. molecules of the solvent (toluene) at room temperature. The probability of such energy transfer was estimated to be 5 X 10'^ s “ ' [13]. For the past 10 years, the methods of nanosecond and picosecond laser spectroscopy permitted the kinetic regularities of EEET to be investigated and the experimental and theoretical results to be compared [14-16]. In some experiments utilizing very high excitation intensities, the interaction of excited states of the donor-acceptor pair was taken into consideration to explain some nonlinear optical effects which were observed under luminescence investiga­ tions. For example, the dependence of the luminescence quantum yield {B) and lifetime (t) of plant objects on the laser pulse intensity [17-20] is explained by the existence of singlet-singlet annihilation and singlet-triplet fusion processes [21,22]. Such effects in which excited state singlets could also interact with one another (singlet-singlet annihilation) were well-documented in organic crystals [23] and in solution [24]. Though the processes with participation of excited states of D and A appear to offer a plausible explanation for the observed dependences of В and т in vivo, the direct experimental investigation of such a phenomenon has not yet been done. In addition the discovered dependences of В and т on the laser intensity may be due to the possible influence of stimulated processes (hght quenching) which must also be taken into account. The observation and study of the EEET regularities with participation of acceptor triplet excited molecules are of interest for serveral reasons. First, from the point of view of the luminescence quenching in concentrated solu- С.Р, Gurinovich et al. / Radiationless intermolecular energy transfer 299 tions of dye and pigment molecules such transfer may be an additional reason for monomer emission quenching, in particular under intense light excitation. Secondly, because of the essential role of triplet molecules in the majority of photochemical reactions, such processes suggest important photochemical im­ plications. For example, it may be a good method for governing the effectivity of some photochemical processes. Knox et al. [25,26] predicted theoretically the possibility of the FEET by inductive resonance interactions between like molecules according to the scheme: ^A**^^A*-l-heat. ( 1) As noted in [26], such a mechanism of singlet-triplet interaction should be effective in concentrated chlorophyll solutions. But the experimental observa­ tion and the study of the regularity of this process in the system of like molecules (homogeneous singlet-triplet-triplet (S-T-T) migration) are very difficult due to the existence of some competing processes (photobleaching, triplet-triplet absorption, quenching of triplet states, etc.). In addition in such systems, it is necessary to taken into account the concentration quenching of luminescence and acceptor triplet molecule formation, which influence greatly the efficiency of homogeneous S -T -T EEET. For example, one- and two-laser excitation procedures were employed by Menzel [27,28] to investigate the possibility of S-T fusion in low concentration solution systems (2X 10“ ^ - 10 M) of chlorophyll “a” and metal-free phthalocyanine separately at 77 К and 293 K. Weak fluorescence quenching has been observed for both systems under investigation (about 8-10%), which has been attributed to the singlet- triplet fusion process. In spite of the fact that the experimental arrangement utilized in [28] was substantially improved, the inquiry of occurrence of S-T fusion by using only the fluorescence intensity measurements was complicated by experimental difficulties caused by some known effects, competitive with S -T -T energy transfer. Besides, some numerical estimations of triplet mole­ cule concentrations in solutions under question and the proposed features of transfer probabihty F for the S -T -T EEET, cited in [27,28] do not permit us to conclude that the observed fluorescence intensity quenching is dominantly caused by singlet-triplet fusion at such low concentrations of molecules. More detailed arguments of our point of view will be given below. In the only work of Bennet [29] known to us, the heterogeneous S -T -T EEET was achieved in a system composed of perylene (С^оп„, = 3 X lO '^M ) and phenanthrene• d|o (Q,,ceptor = 3 X 1 0 “ ^M) in a cellulose acetate film at 77 K. Such heterogeneous transfer has been discovered by a relative decrease in D fluorescence intensity in the presence of triplet-excited molecules of A. But it should be mentioned that the direct comparison of the experimental results obtained with the theoretical calculations was complicated in this case by the influence of trivial absorption of the perylene fluorescence or screeiung the 300 G.P. Gurinovich et al. / Radiationless intermolecular energy transfer perylene excitation by phenanthrene triplet molecules, which would give simi­ lar results and by great phosphorescence of A as well. In spite of the fact that these trivial effects were minimized in [29], their role and influence boundaries were not completely defined, and it would be interesting to carry out the total comparison of the inductive resonance theory and the experimental results. The inductive resonance theory is also directly applicable to the case where the radiative transition D* ^ D + Л;» is spin-forbidden. The triplet-triplet- triplet (T-T-T) energy transfer may be attributed to those processes which are realized by the scheme; ^D*+^A* ^ 'D + ^A **, ^A** ^^A*-b heat. (2) The possibility of such triplet-triplet-triplet migration and triplet-state quenching has been shown by Kellogg [30], who observed that the phosphores- cence-to-fluorescence intensity ratio of 10 M phenanthrene • d ю solutions in cellulose acetate films was dependent on both the sample concentration and the exciting light intensity. Unfortunately, the strict division of the energy transfer process and other known phenomena, which can decrease the phos­ phorescence intensity under cited experimental conditions, was not performed [30], and therefore the results under discussion may be considered as qualita­ tive ones. In addition the Forster-Galanin theory [1,2] does not strictly apply to the case of homogeneous T -T -T EEET due to the fact that concentrations of the excited ^D* and ^A* molecules change simultaneously according to the law of phosphorescence decay. It should be mentioned that most aromatic molecules have overlapping phosphorescence and triplet-triplet absorption spectra, so process (2) is proba­ bly a general phenomenon. For example, the long-range dipole-dipole mecha­ nism of the two triplet-excited molecules interaction (T -T annihilation) was suggested by Parker and Joyce [31] for explanation of P-type delayed fluores­ cence of aromatic molecules in rigid solutions, but direct experimental evidence of existence of such homogeneous T -T -T EEET has not been presented. Besides, for one of the most widely investigated photoactive systems, crystal­ line anthracene, exciton-exciton interactions, including S -T -T EEET, has been reported to result in charge carrier generation [32]. It is the purpose of the present work to attain conditions and to investigate regularities of the homogeneous and heterogeneous EEET with participation of triplet-excited A molecules (schemes (1) and (2)) in concentrated solutions of complex organic molecules. As is seen from process schemes (1) and (2) an additional increase in acceptor emission does not take place in such processes. The discovery of this transfer may be possible only under investigation of changes in the spectral- kinetic characteristics of D (the intensity and lifetime of the emission). The most direct and convincing evidence of existence of the S -T -T or T -T -T EEET is a direct observation of the energy transfer effect on the time G.P. Gurinovich et al. / Radiationless intermolecular energy transfer 301 dependence of the D emission decay following pulsed excitation, because in such experiments all trivial effects are completely excluded. In the present work we use both methods. 2. Homogeneous S -T -T EEET in concentrated solutions of chlorophyll and pheophytin “a” The present series of measurements have been carried out with solutions of chlorophyll “a” and pheophytin “a” in castor oil at 7’=183K , when the solvent is a rigid matrix. Therefore, all diffusion processes, bringing about the deactivation of triplet states, were practically minimized. The experiment was carried out using a standard setup for lamp flash-photolysis. A more detailed description of the experimental setup has been given elsewhere [33,34]. The experimental results for solutions of chlorophyll “a” are given in fig. 1. It is clear by inspection that in dilute pigment solutions (C < 1 X 10“ '*M) relative quantum yields of intersystem crossing and luminescence do not depend on the exciting light intensity. With increasing molecular concentration in solution relative quantum yields increase at 0. Also, the steepness of experimental curves increases with concentration. Analogous dependences were discovered for pheophytin “a” solutions. At the same time the experimen­ tal dependence of the triplet state lifetime on the exciting light intensity is not observed over the whole range of concentrations under investigation (1 X 10 -1 X 10“ ' M). This leads to the conclusion that up to C = 1 X 1 0 “ 'M the bimolecular processes with participation of the triplet states, followed by their deactivation (T-T transfer to an extraneous quencher, T -T annihilation, concentration quenching of triplet states and so on) are not realized in the systems. Therefore, the dependencies of and are stipulated by physical processes connected with an additional deactivation of a singlet excited state. It is interesting to note that these regularities are observed at high con­ centrations when using samples with low optical densities, but at low con­ centrations (C < 10“ ‘*M) such effects are unobservable. So far as the triplet- triplet absorption spectra of chlorophyll and pheophytin are known [35], we can estimate the influence of the triplet-triplet absorption related quenching at different pigment concentrations. It was established in our experiments that this influence and a competition of triplet molecules for incident hght were minimized when using the samples with fixed and low iiutial optical densities in the region of the pigment absorption red band {D = 0.24 for X = 669 nm in chlorophyll solutions). Furthermore, at high pigment concentrations (C > 10 “ ^M) these competitive effects become even smaller (<5% ) since the concentration quenching of fluorescence and triplet state formation decrease the relative triplet concentration in comparison with the diluted solutions [34]. 302 G. P. Gurinovich et al. / Radiationless intemiolecular energy transfer Fig. 1. Dependencies of relative quantum yields of triplet molecule formation (a) and luminescence (b) for chlorophyll “a” solutions in castor oil on pulse photo-excitation intensity at different initial concentrations: 1) 1X10 '‘ M; 2) IX IO ^^M ; 3) IX IO ’ ^M; 4) 5X 10“ ^M; 5) IX IO ^ 'M . r= 183 K. Finally, photobleaching arising from triplet population build-up via the intersystem crossing from the first excited singlet state and conconutant depletion of the ground state population for the flash duration at the constant flash intensity must yield concentration independent fluorescence quenching when the quantum yield of triplet state formation is constant. But if one takes into account a decrease in this yield at high pigment concentrations [34], an increase in fluorescence quenching effect in raising the concentration cannot be fully explained by a photobleaching process. Therefore, it may be concluded that the homogeneous S -T -T energy transfer plays an essential role in the G. P. Gurinooich et at. / Radiationless intermolecular energy transfer 303 fluorescence intensity decreasing with increasing The calculations which will be made below support this conclusion. First of all, there is a good overlap of fluorescence and triplet-triplet absorption spectra for both pigments [35], which is necessary for this type of transfer. Secondly, there exists effective singlet-singlet migration between monomeric molecules in these systems [36]. For estimation of the S -T -T EEET influence on the fluorescence quantum yield we used the modification of the Forster transfer model, taking into account the possibility of excitation migration in a donor molecule system with the following trapping by molecules of A [37]. In accordance to [37] the change of the fluorescence relative quantum yield is expressed by the formula В _ 1 - v^ye'^'[l - erf(y)] “ erf(y)]Дп where (3) 7 = Yd + Ya / ■ + /^ STT^AO Bo is the quantum yield of a solution free of triplet molecules; Cj, and Сд are the concentrations of D and triplet-excited molecules of A; C^o and Сд^^ are the critical concentrations, corresponding with the two types of the transfer D* ^ D and D* -* A; Uq is the parameter which does not depend on the concentration (0 ^ « ^ ^ 1) and defines the quantum losses in migration acts between monomers; еті(у) = is the error integral. Because all spectral-luminescent parameters of chlorophyll “a” have been studied [36,38,39], we analyzed the experimental and theoretical results for that pig­ ment. In accordance with the inductive resonance theory we calculated the values of the critical transfer distance Rq (which is connected with Cq by Cq = [з^гТ^о]” ') from the expression [1,2]: •'о V (4) _ 9 0 0 0 ІП ІО ® 12871-^и“А where v is the wavenumber; is the acceptor decimal extinction coeffi­ cient; f^{v) is the spectral distribution of the donor molecule fluorescence (measured in quanta and normalized for spectral area to 1 in wavenumber scale); N is the Avogadro number; n is the refraction index of the solvent; Bq is the donor fluorescence quantum yield in the absence of transfer; is the orientation factor (A^ = 2 /3 for the isotropic solutions of mean viscosity). Theoretical values of critical distances calculated for S-S and S -T -T EEET in chlorophyll “a” solutions in castor oil are found to be A®® = 54 A [39] and j^STT 42 Д [40]. The triplet-excited molecules concentration at different pulse excitation intensities was measured in T -T bleaching experiments in the main 304 G.P. Gurinovich et а/. / Radiationless intermolecular energy transfer absorption band maximum of chlorophyll “a” (X = 669 nm). For all concentra­ tions under investigation, yield of D luminescence in the absence of triplet molecules was used in calculations with provision for the concentration fluorescence quenching of chlorophyll “a” in castor oil, which had been experimentally investigated in our earlier works [34,36,39]. The experimental data and the calculation results obtained by using formula (3) are presented in table 1. A satisfactory agreement may be expected when «q = 0.92, i.e. the existence of quantum losses in migration acts between monomers is assumed to be valid. These losses are assumed to arise from nonideal calculation of trivial competitive effect contributions to observed luminescence quenching, because for the homogeneous S -T -T EEET these effects can be minimized but cannot be fully excluded. Thus, the existence of the homogeneous S -T -T transfer in concentrated chlorophyll “a” and pheophytin “a” solutions in castor oil at Г = 183 К is the main explanantion for the observed dependence of quantum yields of fluores­ cence and intersystem crossing on the pulse excitation intensity. In this respect the absence of such dependence for concentrated solutions of Na-fluorescein in glycerol at Г = 193 К is due to the low efficiency of the S -T -T transfer in this system because of extremely low concentration of Na-fluorescein triplet-excited molecules at maximum flash excitation intensities caused by low quantum yield of the intersystem crossing in these molecules (y = 0.008) [41]. Our experimental results and calculations permit us to ananlyze some data and estimations in [27,28], where low concentrated (Q ^ < 2 X 1 0 rigid solutions of chlorophyll “a” and metal-free phthalocyanine have been sep­ arately investigated under laser excitation generating low concentration of triplet-excited molecules ( C j= 0.06-0.1 Since the values of critical transfer distance Лп of S -T -T EEET and the theoretical calculations for Table 1 Influence of chlorophyll “a” triplet-excited molecule concentration on fluorescence quantum yield in castor oil at 7"= 183 К c (M) '-A (M) Initial quantum yield Bq (B /S o )"”’ 1X 10“ ’ 0 0.32 1.0 1.0 1X 10“ ’ 1X 10“ ’ 0,32 0.97 1.0 1X 10“ ’ 3X 10“ “ 0.32 0.92 0.97 5X 10“ ’ 0 0.018 1.0 1.0 5X 10“ ’ 3X 10“ “ 0.018 0.50 0.70 5X 10“ ’ 2.5 X 10“ ’ 0.018 0.44 0.50 1X 10“ ' 0 0.002 1.0 1.0 1X 10“ ' 3X 10“ “ 0.002 0.68 0.60 1 X 10“ ' 1X 10“ ’ 0.002 0.62 0.59 G.P. Gurinovich el al. / Radiationless intermolecular energy transfer 305 chlorophyll “a” solutions have not been given in [27], we used for our calculation the value of = 42 A from [40] and estimated an expected efficiency of quenching for conditions cited in [27] by using formula (3). These calculations show the absence of quenching fluorescence effect for Q ^— 2X 10“ ''M and Ct = 2 X 10“ ’ M (B /5o ^0.99). It should be mentioned that the Forster-Galanin treatment of the problem of singlet excitation migration via the dipole-dipole interaction mechanism conformed well with numerous experimental works even in the case of weak fluorescent or nonfluorescent acceptor molecules and in the experiments on sensitized fluorescence when the energy gap AE = E^ ' — E^ ' excludes the back transfer [2,4,5]. Therefore, there are no good grounds for modifying the main equation of the inductive-resonance theory as it has been done in [28]. It is further alleged in [28] that the critical distance for S-T fusion is a property of the molecular system and the irradiation conditions of the particular experi­ mental configuration employed. This point of view is not obvious because it is well known that the critical distance for any energy transfer via inductive resonance interactions is determined only by molecular orientation, fluores­ cence quantum yield of a D molecule, refractive index of the medium and overlap integral of spectra (see eq. (4)). But the efficiency of the S -T -T transfer is certain to depend on the triplet concentration, i.e. on the incident light intensity. Therefore, the energy transfer with participation of triplet- excited molecules may be fully, described by the Forster-Galanin theory without any additonal assumptions. This is well documented in [25,26,29]. Thus, the value of about 200 A for the critical distance found in [28] for metal-free phthalocyanine solutions of low concentrations is surprisingly large. 3. Influence of heterogeneous S -T -T EEET on donor fluorescence intensity For this experiment we developed a procedure which permitted us to investigate completely all trivial processes, to estimate their influence on measured parameters and to isolate the pure transfer phenomenon (fig. 2). The sample fluorescence is excited by a source S (a tungsten lamp 17 V, 100 W with a dc source) through the glass and interference optical filter system F, selecting a required spectral region. The D luminescence is recorded by the experimental setup including an optical filter system Fj, a prism monochromator UM-2, a photomultiplier FEU-38 and a CI-42 storage oscilloscope operating in the driven sweep regime. The A molecules are pumped to a triplet state via a singlet state by light pulses from two IFP-2000 flash lamps ( t,^ 2 — 2 X 10 s) using a light filter system Fj. The samples investigated (thin films) are arranged as shown in fig. 2. Having placed the source S in position В and changing the light filters F| and Fj and the monochromator wavelength in recording we can determine a real concentration of the A triplet-excited molecules at the same 306 G. P. Gurinovich et at. / Radiationless intermolecular energy transfer pulse exciialion of acceptor tr ip le t molecules 1. S is on ; ЛI = ДІ pM. “ ДІ pkosp. 2. S is o ff ; ЛІ = 0 I pkosph. Fig. 2. Schematic of the experimental arrangement for investigation of the dependence of the donor fluorescence intensity on the acceptor triplet-excited molecule concentration. experimental conditions and use the value of Сд for calculations according to the inductive resonance theory. Then, by switching off the source S and pumping the samples by light pulses we can measure directly the A phos­ phorescence (if it exists) which can affect the measurement of the D fluores­ cence change. The primary intention of this stage of the experiments was to reexamine Bennett’s system [29] perylene Ч- phenanthrene but only in polyvinylbuthyral films with the exclusion of trivial effects and direct comparing experimental results with theoretical calculations. The second donor-acceptor pair was the Na-fluorescein-t-chlorophyll “a” system in polyvinylbuthyral films. The elec­ tronic spectra of both systems are shown in fig. 3 and the main spectral- luminescent characteristics and calculated transfer parameters are presented in table!. The values of the orientational factor in eq. (4) were calculated for an isotropic distribution of molecules in a rigid solvent [5]. A relative change of the D luminescence intensity was measured in the experiments when the concentrations of the A triplet molecules in solutions were known. At such conditions, the quenching of D luminescence correlated in time with the deactivation kinetics of the A triplet states. The experimental values of Д/„ were found by extrapolation to an exciting pulse maximum. All measurements were made in a quartz Dewar at 77 K. The low concentration (C < 10“ ^M) ethanolic solutions of the donor-acceptor pairs investigated at 77 К were examined to estimate a direct influence of the exciting light and D fluorescence absorption by A triplet-excited molecules. In these systems the energy transfer G.P. Gurinovich et al. / Radiationless intermolecular energy transfer 307 Fig. 3. Absorption (1) and emission (2) spectra of donor molecules and triplet-triplet absorption spectra (3) of acceptor molecules in polyvinylbuthyral films at 77 K: (a) perylene + phenanthrene; (b) Na-fluorescein +chlorophyll “a”. had been entirely excluded. Furthermore, such examination was carried out with double polyvinylbuthyral films, containing separately D and A molecules at the same concentrations as in mixed solutions. The results obtained per­ mitted us to calculate a true decrease in D fluorescence which was solely due to the transfer process for different optical densities of the triplet-triplet absorp­ tion. The theoretical alternation of a relative fluorescence quantum yield of D in the presence of A triplet-excited molecules was calculated by the formula [1,2]; d x Y (5) where ^ —3.14 for rigid solutions [5]. The comparison of the experimen­ tal and theoretical results are presented in table 3. As one can see from table 3 there is a close conformity between the theory and the experimental studies for Table 2 Spectral-luminescent characteristics and values of S -T -T energy-transfer parameters for the system of fig. 3 Donor-acceptor Refraction Quantum yield of ^ t r i p l e t Overlap ^^thcor pair index of D emission Bq (S') integral (s) (A) solvent n (cm ' M ') (cm "'*M "’) Perylene -1- phenanthrene 1.489 0.94 [45] 27000 0.37X 10"'^ 4.4 39 (v = 20300 cm“ ') [42,43] Na-fluorescein-l-chlorophyll “a” 1.489 0.97 20000 1.44X10“ '^ 2X10"- 3 48 (»- = 20000 cm " ') [44] Table 3 Comparison of experimental and theoretical results of investigation of S-T -T EEET in polyvinylbuthyral films Donor-acceptor C a O a ( B / B o ) ™ “ Influence ( B / B o ) " ' ’ ( B / B o ) ' " ” ' pair (M) (M) (M) of T-T reabsorpt. (%) 0.21 0.42 19 0.52 0.57 Perylene-t- 0.26 0.38 30 0.54 '0.57 phenanthrene 5.2X10^^ 2.3X10” ^ 3.1X10^^ 0.42 0.26 47 0.49 0.57 1.24 0.35 71 0.60 0.57 Na-fluorescein -t- 2X 10“ ^ 1X10” ^ 2.5X10^"' 0.40 0.88 6 0.94 0.92 chlorophyll “a” 2 X 1 0 '^ 5 X 1 0 "’ 5X 10^“ 0.98 0.70 16 0.83 0.84 О Го C> 5oоSi. Si' * Values of acceptor optical densities Од in the singlet state were measured for phenanthrene at X = 394 nm and for chlorophyll “a” at X = 669 nm. G. P. Gurinovich et at. / Radiationless inlermolecular energy transfer 309 the systems. For the perylene + phenanthrene pair this conformity is insensi­ tive to the recording wavelengths (X, =492 nm, X =474 nm, Xj =450 nm), i.e. the D fluorescence quenching effect is not selective in character. Hence we can consider these results as direct evidence of the energy transfer while the reabsorption effect depends on a spectral distribution of a triplet-triplet absorption of A and must change in variation with a recording wavelength. 4. Influence of heterogeneous S -T -T EEET on decay kinetics of donor fluorescence The experiments of this type have been carried out by using the nanosecond kinetic spectroscopy method. For this purpose an experimental setup with separate and simultaneous selective laser pumping of a donor-acceptor pair was used which permitted us to record the D fluorescence decay kinetics in the absence and presence of A triplet-excited molecules. A more detailed descrip­ tion of the experimental apparatus based on a double frequency Q-switched ruby laser has been given in [46]. When exciting the donor-acceptor system by a 347 nm laser pulse ( t,^ 2 ~ 25 ns, E^ = 0.05 J) we measure the D fluorescence lifetime in the absence of A triplet molecules. In exciting simultaneously this system by two laser pulses (X2 = 694 nm and X, = 347 nm) we measured the D fluorescence lifetime in the presence of A triplet molecules pumped to the triplet state by a 694 nm laser pulse ( t,^ 2 ~ 25 ns, £2 ~ 11 J)- In this case the existence of the S -T -T energy transfer can be found by decreasing the D fluorescence lifetime in the presence of the A triplet molecules, and thus one can investigate the regularities and mechanism of such a transfer. All measure- Fig. 4. Absorption (1) and luminescence (3) spectra of pyrene and Sq-S„ (2) and T,-T„ (4) spectra of Mg-phthalocyanine in polyvinylbuthyral films at 293 K. X, =347 nm and Xj =694 nm are the laser excitation wavelengths. 310 G. P, Gurinovich et al. / Radiationless intermolecular energy transfer ments have been carried out at room temperature. The donor-acceptor pair studied is pyrene (D) + Mg-phthalocyanine (A) in polyvinylbuthyral films. The electronic spectra and the main spectral and luminescence characteristics of these molecules are presented in fig. 4 and table 4. The chosen donor-acceptor system satisfies the following conditions: (1) the overlap of D fluorescence and A singlet-singlet absorption spectra is less than the analogous D fluorescence and the same A triplet-triplet absorp­ tion spectra overlap (fig. 4, table 4), i.e. the conditions are secured for a preferable realization of the S -T -T transfer; (2) the probability of A molecule intersystem crossing to the triplet state (гд 10* s “ ') exceeds the deactivation probability of the D singlet-excited state ( /^ =2.6X 10* s “ '), i.e. the A triplet-excited molecule concentration is not equal to zero at the beginning of the D fluorescence emission; (3) the A molecule triplet lifetime (т^ = 1 X 10 “ *s) is considerably greater than the D fluorescence lifetime ( t^ = 3.8 X 10“ ’ s), i.e. the A triplet-excited molecule concentration is constant and not equal to zero during the whole D fluorescence lifetime. The latter condition enables us to use the equations of the inductive resonance theory with no additional assumptions [49]. Both sample constituents have been incorporated into a polyvinylbuthyral Table 4 Spectral-luminescent characteristics and values of energy-transfer parameters of pyrene-Ь Mg- phthalocyanine system in polyvinylbuthyral films (n = 1.489; Cn = lX 10^^M ; Сд = 2.25х 10 Bo =0.72 [47]) Singlet-singlet transfer S* (X = 400 nm) 0.9X 10 ''cm ^' [48] y®® e dv •'o V 2 .7 X 1 0 ~ '''cm ^ ''M ‘ ' ( / j S S ) t h e o r 35 A ( / jS S ) C x p 36 A ( т / т о ) « » “ ' 0.86 ( B / B o ) " P 0.71 ( T / T o ) ' * ' ” 0.84 ( B / B o ) ' * ' ” ' 0.60 Singlet-triplet-triplet transfer 6д (Л = 400 nm) 2X 10 ''c m ' ' M “ ' [48] ^ -r dv •'0 5 .2 X 1 0 ~ '''c m " 'M “ ' ( / j S T T ) t h c o r 40 A ( / j S T T ) e x p 40 A (T/To)-Pb) 0.71 ( T / T o ) '* ’- ' 0.76 Donor fluorescence lifetime % = 440 ns, in the absence of both types of transfer; when singlet-singlet transfer is operating, then tq =380 ns. G. P. Gurinovich el a i / Radiationless intermolecular energy transfer 311 film at the concentrations of pyrene, Cj, = 1 X 10“ ^ M, and Mg- phthalocyanine, Сд = 2.3 X 10“ ^M. The selection of a rigid polymer film as a solvent excluded the possible participation of different diffusion processes in the phenomenon under investigation and decreases essentially the influence of oxygen on the triplet state deactivation. In our investigation special experi­ ments using both laser and lamp flash photolysis show that the laser excitation pumps practically all acceptor molecules (~ 100%) to the triplet state. Conse­ quently, in our calculations we use the value of triplet molecule concentration equal to 2.3 X 10 M. The results of the lifetime measurements are shown in fig. 5. It is evident from fig. 5 that in the presence of A triplet molecules the D fluorescence mean lifetime becomes shorter. Furthermore, noticeable deviations from exponential decay are observed at the initial stage of the time dependence of the D fluorescence decay. It must be noted that the exciting laser pulse duration f « т in order to cause perturbations in the range of interest. Therefore, as was predicted by the Forster-Galanin theory, the observed deviation is due to statistical fluctuations in the distribution of molecules by distances in rigid solutions which lead to the dependence of transfer probability on the time t passing from the excitation instant (i.e. F’d' - a ~ { U) ' )• From fig. 4 it is seen that there are two types of FEET which are observed in the system investi­ gated: (1) S-S transfer to nonexcited singlet A molecules and (2) S -T -T transfer to the triplet-excited A molecules. The homogeneous S-S migration between D molecules may be excluded because of a negligible overlap of the pyrene fluorenscence and absorption spectra. We can therefore apply the equations of the inductive resonance theory [1-5]. The analysis of the theoreti­ cal and experimental results for both types of EEET has been carried out using the formulae: Fig. 5. Time dependence of the pyrene fluorescence decay (C = 1 X 10 ^M) in the absence (1) and in the presence (2) of triplet-excited molecules of Mg-phthalocyanine in polyvinylbuthyral films at 293 K. 312 G.P. Gurinovich et al. / Radiationless intermolecular energy transfer (a) a critical transfer distance (rigid solutions), ,6 90001nl0(/i:20.845)5„ d^ (4a) (b) quenching of the D molecule fluorescence quantum, yield, eq. (5) in this paper; (c) shortening the D fluorescence mean lifetime. 2 ^ 4 i — ? е Ч і 2 r 5 X 1 0 [ 3 3 , 3 4 ] . Separate and simultaneous excitation of D and A molecules has been carried out by using two pulse lamps (IFF-2000, T|^ 2 = 7 X 10 s) with proper filtering of the exciting light and phosphorescence emission. A detailed description of the experimental apparatus and conditions can be found in [52]. Solutions of the compounds under study are placed between quartz glasses with layer thickness / = 3-4 X 10 M. These samples are located in a quartz Dewar at a = 45° to the pulse radiation. The time dependence of the D molecule phosphorescence decay in the absence and presence of A Fig. 6. The normalized absorption (1) and phosphorescence (3) spectra of benzophenone and Sq-S„ (2) and T,-T„ (4) absorption bands of Mg-mesoporphyrin in ethanol at 77 K. A - schematic of the experiment. 314 G.P. Gurirtovich et al. / Radiationless intermolecular energy transfer Table 5 Spectral-luminescent characteristics and values of energy-transfer parameters of benzophenone-l- Mg-mesoporphyrin system in ethanol ( T = l l K; n= 1.3623; Cp = 1.5X 10 M; Sq =0.74 [47,53]) Triplet-singlet transfer *'0 1.2X 10^'’ cm^“ M ' ' 1.6Х10^“ М 4.4Х 10“^ М r'T 0 0 50 А (^ ,TS)CXP 50 А 0.97 0.43 1.0 0.41 Triplet-triplet-triplet transfer •'о V 2.1Х10“ ' А т “ ‘'М ^ ' ^А 0 1.2Х10“ -’ М ^А 1.6Х10“ ''М 3.2Х І0^-'М 55 А (дТТТ)Схр 57 А (Т/То)"’' " 0.96 0.44 (Т/То ) '’'■> 0.95 0.40 Benzophenone phosphorescence lifetime tq =5.3 X 10 s in the absenee of both types of transfer. At Сд = 1.2Х10^^М benzophenone phosphorescence lifetime in the absence of I =4.1 X 10 =* S in accordance with the theory [1.2]. Mg-mesoporphyrininitial triplet state lifetime at 77 К is Tj = 1.1 X 10 ' s. singlet and triplet molecules is shown in fig. 7. The comparison of of a single D (C = 1.5 X 10 ^M) and in a mixture with A in excitation of D molecules only (by one pulse lamp) allows us to assume the existence of the singlet-triplet EEET in the system investigated. Besides, the shortening of the D phosphorescence lifetime in the presence of A triplet-excited molecules serves as a direct illustration of the T -T -T transfer at these conditions. It may be mentioned that deviations from an exponential decay of D phosphorescence are observed in the presence of A molecules for both types of transfer (fig. 7). The physical reason for this effect has been discussed in section 4. The comparison of the theoretical and experimental results has been carried out using eqs. (4-7) in the present work and replacing the D normalized fluores­ cence spectrum by the normalized phosphorescence spectrum in the overlap integral, and taking as a quantum yield of D phosphorescence. The direct experiments with flash photolysis of the donor-acceptor pair investigated shows that under our experimental conditions 75% of A molecules are pumped to the triplet state. Therefore, when T -T -T transfer was investigated [52], the above fact was taken into account in calculation of the ( t/ tq)***®”'^ values and determination of from analysis of D phosphorescence decay curves. The G.P. Gurinovich et at. / Radiationless intermolecular energy transfer 315 1 phosplt. Fig. 7. Observed decay curves for the system of fig. 6 : 1 - pure benzophenone (Cd = 1.5X10“ ^M); 2 - benzophenone and singlet molecules of Mg-mesoporphyrin (Сд =4.4X 10 M); 3 - benzophe­ none and triplet-excited molecules of Mg-mesoprophyrin (Сд =3.2X 10“ ^M) T = l l K. numerical results are tabulated in table 5. From analysis of the data obtained we conclude that there is good agreement between the experimental results and the calculations according to the inductive resonance theory. At the same time a decrease in pulse excitation intensity or dilution of D and A concentrated solutions up to Cjj = 1.4 X 10~^M and С д = 4 .2 Х 1 0 “ ^М results in the disappearance of the dependence of the Ц phosphorescence decay on the presence of A triplet molecules. Thus, it is evident from the above that the efficiency of the observed process depends essentially on the A triplet-excited molecule concentration. 6. Conclusion Homogeneous and heterogeneous long-range radiationless transfer with participation of triplet-excited acceptor molecules leads to reduction of quan­ tum yield (sects. 2 and 3) and donor emission lifetime (sects. 4 and 5) with no additional change of acceptor emission. It has been directly shown that the energy transfer from donor excited singlet or triplet levels to acceptor excited triplet levels in a rigid condensed medium can be reasonably explained by the Forster-Galanin dipole-dipole inductive-resonance mechanism. Supporting evidence is given in sects. 4 and 5 of this paper where a direct experimental and theoretical verification has been carried out by investigation of time 316 G. P. Gurinovich et al. / Radiationless intermolecular energy transfer dependence of the donor emission decay following a pulse excitation. Such a transfer can take place even in the case where the radiative transition in donor molecules is spin-forbidden. Even though the donor dipole transition moment is of extremely small magnitude, a long-range radiationless transfer is signifi­ cant in donor de-excitation due to the equally forbidden nature of all compet­ ing processes. The observed phenomenon must take place in all molecular systems excited by an intense light flux and also in the case of an effective energy migration where the meeting of two or more excitons can lead to their interaction (high concentrated solutions, molecular crystals, photosynthetic apparatus of plants and so on). References [1] Th. Forster, Disc. Farad. Soc. 27 (1959) 7. [2] M.D. Galanin, Trudy phys. instituta im. P.N. Lebedeva. Acad. Nauk SSSR, 12 (1960) 3. [3] D.L. Dexter, J. Chem. Phys. 21 (1953) 836. [4] V.L. Ermolaev. E.N. Bodunov, E.B. Sveshnikova and T.A. Shakhverdov, Besizluchatelnii perenos energii electronnogo vozbuzhdeniya (Nauka, Leningrad, 1977). [5] V.M. Agranovich and M.D. Galanin, Perenos energii electronnogo vozbuzhdeniya v con- densirovannih sredah (Nauka, Moscow, 1978). [6] M.V. Alfimov, LG. Batekha and V.A. Smirnov, DAN SSSR 185 (1969) 626. [7] V.L Gerko, Yu.B. Shekk, LG. Batekha and M.V. Alfimov, Opt. i spektrosk. 30 (1971) 456. [8] V.L Gerko, L.S. Popov and M.V. Alfimov, Spectrosc. Lett. 11 (1978) 479. [9] V.A. Benderskii, V.Kh. Brikenstein, A.G. Lavrushko, P.G. Filippov and V.A. Yatsenko, Chem. Phys. Letts. 56 (1978) 443. [10] J. Kaplan and J. Jortner, Chem. Phys. Letts. 52 (1977) 202. [11] J. Kaplan and J. Jortner, Chem. Phys. 32 (1978) 381. [12] V.L. Ermolaev. A.A. Krasheninnikov and A.V. Shablya, DAN SSSR 248 (1979) 389. [13] V.L. Ermolaev, A.A. Krasheninnikov, V.A. Lubimtsev and A.V. Shablya, Izv. Acad. Nauk SSSR, Ser. Phys. .44 (1980) 709. [14] D. Rehm and K.B. Eisental, Chem. Phys. Letts. 9 (1971) 387. [15] R.W. Anderson, R.M. Hochstrasser, H, Lutz and G.W. Scott, Chem. Phys. Letts. 32 (1975) 204. [16] G. Porter and C.J. Tredwell, Chem. Phys. Letts. 56 (1978) 278. [17] D. Mauzerall, Biophys. J. 16 (1976) 87. [18] A.J. Campilo, S.L. Shapiro, V.H. Kolman, K.R. Winn and K.C. Hyer, Biophys. J. 16 (1976) 93. [19] M. Seibert, R.R. Alfano and S.L. Shapiro, Biochim. Biophys. Acta 292 (1973) 493. [20] G.S. Beddard, G. Porter, C.J. Tredwell and J. Barber, Nature 258 (1975) 166. [21] A.J. Campilo, V.H. Kolman and S.L. Shapiro, Science 193 (1976) 227. [22] A.J. Campilo and S.L. Shapiro, in: “Topics in Applied Physics”, V. 18. Ultrashort Light Pulses, Picosecond Techniques and Applications, ed. S.L. Shapiro (Springer-Verlag, Berlin, 1977) p. 317. [23] A. Bergman, M. Levin and J. Jortner, Phys. Rev. Letts. 18 (1967) 593. [24] R.M. Hochstransser and A.C. Nelson, Opt, Commun. 18 (1976) 361. [25] T.S. Rahman and R.S. Knox, Phys. Stat. Sol. (b) 58 (1973) 715, [26] R.S. Knox and V.J. Ghosh. Photochem. Photobiol. 22 (1975) 149. [27] E.R. Menzel, Chem. Phys, Letts. 26 (1974) 45. G.P. Gurinovich et al. / Radiationless intermolecular energy transfer 317 [28] E.R. Menzel, J. Chem. Phys. 66 (1977) 67. [29] R.G. Bennet, J. Chem. Phys. 41 (1964) 3048. [30] R.E. Kellog, J. Chem. Phys. 41 (1964) 3046. [31] C.A. Parker and T.A. Joyce, Chem. Commun. (1967) 744. [32] C.R. Johnston and L.E. Lyons, Aust. J. Chem. 23 (1970) 1571. [33] A.P. Losev, E.I. Zen’kevich and E.I. Sagun, Zh. Prikl, Spektrosk. 27 (1977) 244. [34] A.P. Losev, E.I. Zen’kevich and E.I. Sagun. Izv. Acad. Nauk SSSR, Ser. Phys. 44 (1980) 783. [35] G.P. Gurinovich, A.N. Sevchenko and K.N. Solov’ev, Spectroscopiya chlorophylla i rodstven- nyh soedinenii (“Nauka i Tekhnika”, Minsk, 1968) p. 371. [36] E.I. Zen’kevich, A.P. Losev and G.P. Gurinovich, Migratsiya energii electronnogo vozbuzh- deniya mezhdu pigmentami photosyntheticheskogo apparata rastenii (Preprint Insituta Physiki Acad. Nauk BSSR, Minsk, 1971) p. 60. [37] C. Bojarski, J. Kusba and G. Obermuller, Z. Naturforseh. 26a (1971) 255. [38] S. Claesson, L. Lindquist, and B. Holmstrom, Nature 183 (1959) 601. [39] G.P. Gurinovich, A.P. Losev and E.I. Zen’kevich, Spectr. Letts. 11 (1978) 493. [40] G.P. Gurinovich, E.I. Zen’kevich and E.I. Sagun, Izv. Acad. Nauk SSSR, Ser. Phys. 44 (1980) 695. [41] E.I. Zenkevich, E.I. Sagun, A.P. Losev and G.P. Gurinovich, Zh. Prikl. Spektrosk. 32 (1980) 1047. [42] R.A. Keller and S.G. Hadley, J. Chem. Phys. 42 (1965) 2382. [43] D. Lavalette, C. R. Acad. Sci. Paris В 266 (1968) 279. [44] H. Linschitz and K. Sarkanen, J. Amer. Chem. Soc. 80 (1958) 4826. [45] J.B. Birks, Photophysics of Aromatic Molecules (New York, 1970). [46] K.I. Salokhiddinov, B.M. Dzhagarov and I.M. Bitewa, Opt. i Spektrosk, 47 (1979) 381. [47] C.A. Parker, Photoluminestsentsiya rastvorov (“Mir”, Moskwa, 1972). [48] M.P. Tsvirko, V.V. Sapunov and K.N. Solov’ev. Opt. i Spektrosk. 34 (1973) 1094. [49] G.P. Gurinovich, E.I. Zen’kevich and E.I. Sagun, Doklady Acad. Nauk BSSR 24 (1980) 315. [50] J. McVie, R.S. Sinclair and T.G. Truskott, J. Chem. Soc. Farad. Trans. II, 10 (1978) 1870. [51] V.L. Ermolaev and E.B. Sveshnikova, Izv. Acad. Nauk SSSR, Ser, Phys. 26 (1962) 29. [52] G.P. Gurinovich, E.I. Zen’kevich and E.I. Sagun, DAN SSSR 257 (1981) N. 1. [53] E.H. Gilmore, G.E. Gibson and D.S. McClure, J. Chem. Phys. 20 (1952) 829.