6620
J . Phys. Chem. 1988, 92, 6620-6625
Solvent and Substituent Effects on Intermolecular Electronic Energy Transfer between Saturated Amines and Benzene: Evidence of Exciplex Intermediates Arthur M. Halpern* and Stephen L. Frye Department of Chemistry, Northeastern University, Boston, Massachusetts 021 15 (Received: April 1, 1988)
Intermolecular electronic energy transfer is studied in the system N,N-diethylmethylamine (DEMA) and benzene (B) in tetrahydrofuran (THF) solution at 298 K. The results are compared with the same system in n-hexane solution. In THF, sensitized DEMA fluorescence (340 nm) is observed consequent to excitation of B. Examination of the fluorescencespectrum of the mixture relative to those of the DEMA and B components reveals emission with a maximum at ca. 380 nm; this is assigned as exciplex fluorescence. The photokinetics are consistent with this interpretation; when B is photoexcited (276 nm), fluorescencedecay, monitored at 340 nm (DEMA), manifests three components. A kinetic scheme is developed involving the exciplex, that is irreversibly formed from photoexcited benzene but reversibly formed from photoexcited amine. The different properties of the DEMA/B system in THF via-%-visn-hexane are associated with the strong solvation effects of the DEMA* Rydberg state in THF. Exciplex formation from DEMA* is much less efficient relative to B*, and this is interpreted in terms of the decreased mobility of the solvated photoexcited amine and possibly increased steric hindrance posed by the planar DEMA* molecule. On the basis of the fact that only exciplex emission and not energy transfer is observed in an intramolecular system, it is concluded that exciplex formation demands more stringent geometrical requirements than does energy transfer. Fluorescence spectra and photokinetic schemes are described for DEMA/fluorobenzene and 1,4-difluorobenzene in n-hexane and THF solution.
Introduction Examples of efficient electronic energy transfer that takes place in fluid media between species possessing a small overlap factor (the spectral overlap between the emission spectrum of the donor and the absorption spectrum of the acceptor) are uncommon. It is important, therefore, to characterize such a system when it is encountered and to elucidate its mechanism. An example of collisional energy transfer beween poorly overlapping species was reported several years ago by Halpern and Wryzykowska’ in a study of the standard tertiary amine N,N-diethylmethylamine (DEMA) and benzene (B) in n-nexane solution at ambient temperature. For this system, the overlap between DEMA emission and benzene absorption is very small. This overlap is also small with respect to B emission and DEMA absorption. For example, DEMA absorbs significantly only below ca. 240 nm, exhibits the onset of fluorescence at 265 nm, and reaches a maximum in intensity at 287 nm, respectively, while B absorbs below 263 nm and reaches an emission maximum at ca. 278 nm. In fact, the fluorescence spectra of both benzene and DEMA strongly overlap in n-hexane solution. Notwithstanding the poor DEMA fluorescence/benzene absorption overlap, selective irradiation of DEMA (at 220 nm) in the presence of benzene results in the efficient sensitization of benzene fluorescence. Furthermore, irradiation of benzene (265 nm) in a system with a sufficiently high DEMA concentration produces sensitized DEMA fluorescence. A detailed photokinetic investigation of this system’ led to the conclusions that (1) energy transfer between excited DEMA and benzene is diffusion controlled and (2) energy transfer in the reverse direction occurs with an efficiency of ca. 0.13. It should be noted that the very poor spectral overlap between the amine and benzene precludes long-range dipole-dipole (Forster) transfer. The Forster distance is calculated to be only a few angstroms. In view of the fact that a significant change in geometry occurs between DEMA in the ground state (pyramidal) and its electronically excited state (planar),2 the sensitization of the amine by electronically excited benzene would appear to require considerable excitation of the out-of-plane bending motion in the amine, Le., a distortion of the C-N-C bond angles. Entry into the electronically excited-state DEMA potential energy surface directly via the pyramidal configuration would occur at about 5000 cm-’ above the thermally equilibrated levels of S1. Based, in part, on this evaluation, the mechanism of energy transfer was postu(1) Halpern, A. M.; Wryzykowska, K. Chem. Phys. Lett. 1981,77,82. (2) See: Halpern, A. M.; Ondrechen, M. J.; Ziegler, L. D. J . Am. Chem. SOC.1986,108, 3901.
0022-3654/88/2092-6620$01.50/0
lated to involve the intermediacy of an excited complex between DEMA and benzene, Le., an exciplex. N o direct evidence for such a species was found, however, possibly because of its low emission yield and/or significant overlap between the exciplex emission and that of both DEMA and benzene fluorescence. Working at much higher benzene and amine (triethylamine) concentrations in cyclohexane (0.56 and 0.1 M, respectively), Mattay and Leismann observed emission at 345 nm that they ascribed to the amine-benzene exciple^.^ Beecroft and Davidson also observed such emission, at 350 nm, in cyclohexane solution at high amine and benzene concentrations that they assigned as an amine-benzene exciplex stabilized by additional benzene molecule^.^ A subsequent photokinetic study of this system revealed complex fluorescence decay at 280 and 345 nm consisting, respectively, of a sum and difference of two components (0.47 and 4.9 n ~ ) .Photochemical ~ studies involving the amine-mediated photoreduction of benzene and its derivatives led Bryce-Smith and wworkers to postulate that in nonpolar solution an amine-benzene complex having appreciable polar character is formed via electron transfer from the amine to photoexcited benzene.6 In such an amine-benzene exciplex, charge transfer from the amine to the benzene can be presumed to distort the amine geometry,6 increasing the C-N-C bond angles (consistent with the planar structure of the amine radical cation), and thus the sensitization of S1 would be facilitated because of enhanced Franck-Condon factors. In essence, then, the apparently “nonvertical” nature of the benzene-to-DEMA energy transfer would be clarified by the role of the exciplex. The direct and exciplex-mediated energy-transfer steps are summarized in reactions 1 and 2, where A and B represent amine and benzene, B* B*
+A
+ A - B + A* 4
(B-A)*
-
B
(1)
+ A*
(2)
and reaction 2 occurs with higher probability relative to (1). An intramolecular exciplex between the phenyl and N,N-dialkylamino moieties has been observed and studied, both in the (3) Mattay, J.; Leismann, H. Tetrahedron Lett. 1978,44, 4265. (4) (a) Beecroft, R. A.; Davidson, R. S.; Whelan, T. D. J. Chem. Soc., Chem. Commun. 1978,911.(b) Beecroft, R.A,; Davidson, R. S . Chem. Phys. Letf. 1981,77,77. (5) Liesmann, H.; Scharf, H.-D.; Strassburger, W. S.; Wolmer, A. J. Photochem. 1983,21, 215. (6)(a) Bryce-Smith, D.; Gilbert, A.; Kiunkiin, G. J . Chem. SOC.,Chem. Commun. 1973,330. (b) Bellas, M.; Bryce-Smith, D.; Clarke, M. T.; Gilbert, A,; Klunklin, G.; Krestonosich, S.; Manning, C.; Wilson, S . J . Chem. SOC., Perkin Trans. I1977. 2511.
0 1988 American Chemical Society
Evidence of Exciplex Intermediates vapor phase' and in solvents of varying polarity? Evidently, under favorable circumstances, emission from the bound excited electronic state formed from the association of these two moieties can be observed. The existence and thus role of the exciplex as an intermediate in intermolecular energy thus appear plausible. The ability to observe emission from an exciplex intermediate is, of course, a kinetic as well as thermodynamic issue and thus depends on the magnitudes of its formation, radiative, and other dissipative rate constants. To explore further the mechanism of energy transfer between amines and benzene (and its derivatives), we have investigated the DEMA/benzene system in a relatively polar solvent, tetrahydrofuran (THF). The strategy is to stabilize the polar exciplex and thus to shift the emission of this species to lower energies, i.e., away from the interfering fluorescence of donor and acceptor. Tetrahydrofuran is perhaps the optimal polar solvent medium for this study because the fluorescence of DEMA (and other tertiary amines) is severely quenched in more polar solvents (e.g., acet ~ n i t r i l e ) . ~Moreover, because the fluorescence lifetime and spectra of saturated tertiary amines in this solvent are dramatically different from those in nonpolar solution, photophysical studies involving amines acquire a methodological advantage. lo These properties will be discussed below. To provide consistency with the earlier studies, we used DEMA as the saturated amine species. However, to gain further insight into the amine/benzene system, we also examined three aromatic partners of varying electron-accepting ability: fluorobenzene, 1,Cdifluorobenzene, and 1,3,5-trifluorobenzene. The photophysical properties of these benzene derivatives and DEMA were also studied in n-hexane solution.
Experimental Section DEMA (Aldrich) was refluxed and distilled over CaH2 immediately prior to use; benzene (Burdick and Jackson) and the fluorinated derivatives (Aldrich) were distilled before use. Materials were stored and manipulated in a N2-purged drybox; the DEMA/aromatic mixtures were prepared just prior to the experiments. Solvents (n-hexane and THF, Burdick and Jackson spectroscopic grade) were dried over lithium aluminum hydride, distilled, and stored in a drybox. Samples were degassed via at least four freeze-pump-thaw cycles. Absorption spectra were obtained by using a Cary 14 or Varian 2300 spectrophotometer. Fluorescence spectra were obtained with a homemade dc fluorimeter; the excitation source was a 60-W D2 lamp. Excitation and emission bandpass values were 3.2 and 2.0 nm, respectively. Fluorescence decay data were obtained by using a time-resolved single-photon-counting apparatus. Excitation was achieved with a thyratron-gated D2 (0.5 atm) flash lamp running at 30 kHz. The light was dispersed with a 0.25-m Jarrell-Ash monochromator; the bandpass was 1.6 nm for most experiments, otherwise it was 3.2 nm. Emission was isolated by using interference filters (Corion) centered near the maximum intensity (bandpass, typically 10 nm). The sample was mounted in a thermoelectrically controlled constant-temperature block. The temperature 296 K, was stable to 0.1 K. Reconvolution analysis was performed on a dedicated LSI-11/23 microcomputer using an iterative procedure capable of analyzing up to three exponential components." x 2 values were typically 1.2-1.5. The assignment and confirmation of certain rate constants was achieved by curve fitting using EUREKA.
Results and Discussion Absorption and Fluorescence Spectra. There is little change in the absorption and fluorescence spectra of the aromatic mol(7) Halpern, A. M. J. Am. Chem. SOC.1984, 106, 6484. (8) (a) Vanderauweraer, M.; De Schryver, F. C . Bull. SOC.Chim. Belg. 1979,88, 227. (b) Swinnen, A. M.; Van Der Auweraer, M.; De Schryver,
F. C.; Windels, C.; Goedeseek, R.; Vawerem, A,; Meeus, F. Chem. Phys. Lett. 1983, 95, 461.
(9) Halpern, A. M.; Wryzykowska, K. J . Photochem. 1981, 15, 147. (10) Halpern, A. M. J. Phys. Chem. 1981, 85, 1682. ( 1 1) Frye, S. L.; KO,J.; Halpern, A. M. Photochem. Photobiol. 1984,40, 555.
The Journal of Physical Chemistry, Vol. 92, No. 23, 1988 6621 WAVELENGTH,
NM
1 O3 WAVENUMBERS Figure 1. Absorption and fluorescence spectra of DEMA and B (separately) in T H F solution at 298 K.
> I n
I
A
\ 260
300
340
380
420
460
500
WAVELENGTH, NM
Figure 2. Fluorescence spectrum of a mixed solution of DEMA (4.1 X M) and B (5.6 X lo-' M) in T H F a t 298 K, excited a t 267 nm. Also shown is the difference spectrum between the DEMA/B mixture and a pure DEMA fluorescence spectrum, normalized at the high-energy end.
ecules in T H F solution relative to n-hexane. Each of these compounds manifests the first strong vibronic band at ca. 254 nm (0 near 265 nm) and a minimum in the absorption spectrum a t ca. 225 nm. The latter characteristic is important because it provides a means for relatively selective excitation of DEMA in the presence of an aromatic species. As in n-hexane, the absorption spectra of the aromatic molecules in THF become more diffuse with increasing fluorine substitution. Similar to the above observations, the fluorescence spectra of these molecules do not manifest appreciable solvent polarity effects in THF relative to n-hexane. The Franck-Condon fluorescence maxima of the aromatic compounds are observed near 280 nm. In THF solution, the absorption spectrum of DEMA is structureless, as it is in n-hexane, and commences at ca. 250 nm. There is a slight blue shift in THF relative to the nonpolar solvent, so that only end absorption is observed (to 200 nm). DEMA, like other saturated tertiary amines, however, manifests a dramatic medium effect with respect to its emission properties. For example, in THF, fluorescence is considerably red shifted and is much broader relative to n-hexane solution, having a maximum at 340 nm (cf. 280 nm in n-hexane). Figure 1 portrays the absorption and emission spectra of THF solutions of benzene and DEMA. Moreover, in T H F the fluorescence quantum efficiency of DEMA is reduced to about 22% of that in n-hexane (in which it is 0.64); nevertheless, the lifetime is longer, 50 ns, as compared with 25 ns in n-hexane (in the low-concentration limit). The considerable decrease in the radiative rate constant indicated by these data has been noted before and may be attributed to the strcng solvation of the 3s Rydberg excited state of DEMA in THF.'O Presumably, the penetration of the 3s-like electron into the higher dielectric medium and the consequent stabilization that this achieves12 reduce the overlap between the soluated excited state (nN,3s) and the more compact ground state. The immediate consequence of this interesting solvent effect on the energy-transfer (12) Muto, Y.; Nakato, 591.
Y.; Tsubomura, H. Chem. Phys. Lett. 1971, 9,
6622 The Journal of Physical Chemistry, Vol. 92, No. 23, 1988
Halpern and Frye 1o3
I
,
/
1 o2 2
u
z Q
I U
In
\
-
I
/
5 0 U
IC
-I 0
050
100
150
200
I02[DEMAI
1
Figure 3. Plot of l/ie(see text) versus DEMA concentration in THF solution at 298 K. Excitation wavelength is 267 nm.
study discussed here is that the emission of the amine is now shifted to lower energy relative to the aromatic species. Thus, the intensity and dynamics of the amine emission can be followed separately from those of the aromatic partners. We first discuss in detail the prototype system, DEMA/benzene, in T H F solution. Steady-State Results. Excitation of a mixture of DEMA and benzene in deaerated THF solution (4.1 X and 5.6 X M, respectively) at 267 nm, a wavelength at which radiation is nearly exclusively absorbed by benzene, produces an emission spectrum shown in Figure 2. This spectrum clearly reveals emission from both benzene and DEMA species at 276 and 340 nm, respectively, and thus provides direct evidence for energy transfer from benzene to DEMA. When excited at 267 nm, a DEMA blank of equal concentration produced negligible emission. Further analysis of the spectrum shown in Figure 2 relative to separate DEMA and benzene blanks reveals additional emission intensity at longer wavelengths. The difference spectrum, obtained by subtracting the maximum-normalized spectra of the blanks = 267 nm) from that (DEMA, A,, = 230 nm; benzene, A, 380 observed in the mixture, produces a broad band with A, nm; this is also shown in Figure 2. This feature is assigned as emission from the DEMA/benzene exciplex in THF. Additional support for this assignment is provided by the photokinetic data discussed below. Intramolecular exciplex emission from 1phenyl-3-(N,N-dimethylamino)propane was reported by De Schryver et al. to occur at ca. 385 nm in T H F solution.8 When the same solution (see above) is excited at 225 nm, a position at which light is principally, although not exclusively, absorbed by DEMA, the emission produced appears at first inspection to resemble that of the DEMA blank, Le., no benzene fluorescence can be discerned. Nevertheless, subtraction of the DEMA blank fluorescence (normalized at 340 nm) from this spectrum also produces the weak broad (exciplex) band at ca. 380 nm observed with A,, = 267 nm. Excitation of DEMA/benzene mixtures containing higher benzene concentrations (Le., >5.6 X I 0-3 M) results in the appearance of structured emission near 276 nm. This is benzene emission probably produced via direct excitation at 225 nm, although the role of benzene emission sensitized by DEMA cannot be excluded. Photokinetics. Fluorescence produced with an excitation wavelength of 267 nm and monitored a t 287 nm (see Figure 1) produces single-exponentialdecay. The lifetime of this component, referred to as 7 B 9 is associated with the decay of photoexcited benzene and is dependent only on the concentration of DEMA; M, 78 = 9.8 ns. A plot for example, for [DEMA] = 8.0 X of sB-]versus [DEMA] is linear, having a slope of 7.4 X lo9 M-I s-’; see Figure 3. The intercept, 5.0 X lo7 s-l, is in agreement with the lifetime of a benzene blank in T H F (5.6 X M). The
-
u TIME
Figure 4. Fluorescence decay curve of a mixture of DEMA (3.0 X M) and B (5.6 X lO-’ M) in THF excited at 267 nm and analyzed at 340 nm. The reconvoluted function indicated in eq 3 is superimposed.
significance of these data will be discussed below. When fluorescence is monitored at 340 nm, the maximum of the sensitized DEMA emission, a complex decay curve, shown in Figure 4, requiring three components is reproducibly and systematically observed. Thus for a solution in which benzene and 3.0 X M, and DEMA concentrations are 5.6 X respectively, the following decay law, obtained from reconvolution analysis (see the Experimental Section), is observed: Z&t)
= exp(-t/35)
+ 0.32 exp(-t/5.5)
- exp(-t/l5)
(3)
where t is in nanoseconds. It is noteworthy that the intermediate-lived component of 15 ns matches the observed value of T~ for the same solution when emission is analyzed at 287 nm. This agreement, or consistency, in lifetimes was observed throughout the ranges of benzene and DEMA concentrations studied. The longest lived component, e.g., 35 ns in the above benzene/DEMA system, referred to as 7A, is associated with the quenched DEMA fluorescence. The value of 7 A was found to be dependent on both the benzene and DEMA concentrations. Stern-Volmer plots of 7A-I versus [B] (at fKed [DEMA]), are observed to be linear linear and furnish intercepts consistent with the lifetimes of DEMA blanks at the respective concentration studied. These intercept (or DEMA blank) data reflect DEMA self-quenching in T H F solution (vide infra). Several plots are shown in Figure 5. The slope of 7A-I versus [B], which is approximately invariant with M has a value respect to [DEMA] in the range (4-17) X of 5 X lo8 M-‘ s-’. The value OF the short-lived component, e.g., 5.5 ns in the decay law cited above and referred to as 7E,is independent of both benzene and DEMA concentration under the experimental conditions indicated. With respect to excitation at 225 nm, at which point DEMA is the principal light absorber, the fluorescence intensity at 287 nm is very weak and can be attributed to direct excitation of benzene. When analyzed at 340 nm, however, where the emission is much stronger, the fluorescence reveals double-exponential decay in which one component has a negatiue amplitude. For example, a solution in which [B] = 5.6 X M and [DEMA] = 3.0 X M manifests the following decay law determined from reconvolution analysis: I f ( ? ) = exp(-t/35) - 0.9 exp(-t/5.5)
(4)
and demonstrates that the two lifetimes obtained are in good agreement with the longest and shortest lifetimes observed in the same sample in which A,, = 267 nm and Acmiss = 340 nm; see eq 3.
The Journal of Physical Chemistry, Vol. 92, No. 23, 1988 6623
Evidence of Exciplex Intermediates
TABLE I: Rate Constants Associated with DEMA (A) and Benzene (B) in THF Solution Analyzed According to Scheme I kA, S-' k q ~M-' , S-' kM. M-' s-l kAE, s-' kE, S-I kE, M-' S-' 2.0 x 107 1.1 x 109 5.0 X lo8 1 x 17 3.9 x 108 7.4 x 109
L
--NgA ' . __,_I.. /
I\
..k..
k,, s-' 5.0 x 107
E'
A
Figure 6. Schematic energy level diagram for DEMA and benzene in THF solution. The circle enscribing the geometrically relaxed excited amine denotes a solvation shell. SCHEME I kEA[el
A*=
A
IkA
A 20 0
0.50
100
150
200
1 02[B1
Figure 5. Plots of l/rA(see text) versus benzene concentration in THF at 298 K for various DEMA concentrations: (A) 4.0 X (B) 8.0 X 10-3; (c) 1.2 x 10-2; (D)1.7 x 10-2 M.
The photokinetic data presented thus far demonstrate that in THF solution, energy transfer takes place between photoexcited benzene and DEMA and that, moreover, an emissive exciplex appears to play a role in this process. This interpretation is based on the following summary observations: (1) sensitized DEMA emission is observed as a result of benzene excitation; ( 2 ) three different components, manifesting (at least) three different species, are observed in the decay curves; (3) negative amplitudes are observed for the two shorter lived components. The data also suggest that reverse transfer, Le., sensitized benzene emission consequent to DEMA excitation, does not take place in THF solution. That is, sensitized benzene emission is not observed for A,, = 225 nm; moreover, a decay component possessing the characteristic lifetime of benzene is likewise not observed at this excitation wavelength. It should be mentioned that in n-hexane solution, energy transfer from DEMA to benzene takes place at the diffusion-controlled rate. This dramatic reversal in the propensity of energy transfer is the result of the significant lowering of the energy of the emissive state in DEMA in T H F solution relative to n-hexane.', This interpretation is summarized in a semiquantitative energy level diagram shown in Figure 6 . For simplicity, benzene is represented by a two-level scheme, while DEMA is depicted with four levels: the geometrically relaxed ground (pyramidal) and excited (planar) states, as well as the Franck-Condon excited and ground states (the states reached directly via absorption and emission, respectively). The intermediate species, assigned as the exciplex, is situated qualitatively in this diagram because, while the emission energy of this species is known, the contribution of binding and repulsion energies has not been determined. On the basis of the nature of the decay curves observed (and presented above) under various excitation and emission conditions, we assume that dissociation of the exciplex into excited benzene and ground-state amine is slow (viz. benzene fluorescence decay is single exponential). Hence, the energy of the exciplex is presumed to lie considerably below the (local) excited state of benzene. A photokinetic scheme that is consistent with the observed decay curves is shown in Scheme I, where B, A, and E* represent benzene, DEMA, and the exciplex. Subsequent to &pulse pulse
k [AI
E*A BO-
IE
(AB)!
hve
B
/rB
B
excitation of benzene, Scheme I predicts the followi g rate laws for the three species, B*, A*, and E*: [B*](t) = [B*], exp(-Xt) (5) MIB*lCJkAE [(L+- L-) exp(-Xt) = (X- L-)(X - L+)(L+- L-) (X- L+) exp(-L-t) - ( X - L-) exp(-L+t)] [A*], [ ( Y- L-) exp(-L-t) - ( Y - L+) exp(-L+t)] ( 6 ) (L+ - L-)
+
+
M[B*l,
[ ( Z - X)(L+ - L-) exp(-Xt)
(X- L J ( X - t+)(L+- L-) ( Z - L ) ( X - t+)exp(-L-t) - ( Z - t+)(X- L-) X
where
+
X = kB + ~ E B [ A ] ;Y = kE + k,; =
kA
+ kQAIAl + kEAIBl
M = ~ E B [ A ] ;N = ~ E A [ B ] 2L+ = ( Y + z ) f ( ( Y + z ) z - 4 ( Y z - N k ~ ~ ) ) " 2 The rate constant k,, accounts for the self-quenching of DEMA; see the intercepts of the plots in Figure 5 . The decay laws presented above for the excited amine and the exciplex are separated into terms that denote the separate decay functions with respect to initially excited amine [A*], and benzene [B*], concentrations. The decay law for excited amine, for example, predicts that for [B*], >> [A*],, Le., A,, = 267 nm, triple-exponential decay should result in which one component is negative; this is observed (see eq 3). Moreover, for [A*], >> [B*],, Le., A,, = 225 nm, double-component decay should be produced, one of which has a negative component; this is also observed (see eq 4). Also, on the basis of a phenomenological comparison of the observed fluorescence decay curves with eq 5-7, the decay pa, T E can be assigned to the quantities L-, X, rameters T ~ T, ~ and and L+, respectively. Model calculations using the fluorescence decay laws (5)-(7) to fit experimental data (e.g., Figures 3-5 and eq 3 and 4) allow assignments to be made of the constants of Scheme I. These results are summarized in Table I. It is noteworthy that exciplex formation appears to be significantly less efficient between excited amine and benzene (kEA) than between excited benzene and amine (kEB). In the latter case,
6624
The Journal of Physical Chemistry, Vol. 92, No. 23, 1988
SCHEME I1 ~EBIAI
TABLE 11: Rate Constants for the Fluorine-Substituted Benzenes/DEMA System as Analyzed According to Scheme 11"
kFAlB1
E*-
FBI-
Halpern and Frye
A*
kEB. EA, k g , s-' M-' sd kE, s-' M-' s-I k A, s-I B 0.50 0.74 3.9 0.05 0.20 FB n-hexane 13.0 1.7 1.2 2.9 0.38 FB THF 13.0 2.4 0.5 1.3 0.20 DFB n-Hexane 2.0 1.3 0.9 2.8 0.38 DFB THF 2.0 1.0 0.2 2.0 0.20
B
t
4
t
FB
(FB+A)'
A
-
exciplex formation is nearly diffusion controlled (k,fl 1 X 1O'O M-I s-l, on the basis of the Stokes-Einstein-Smoluchowski equation). The fact that k E A is ca. 7% that of kEBmay be a consequence of the more profound solvation of A* vis-8-vis B* by THF, thus increasing the activation barrier to exciplex formation. The evidence of dramatic solvation effects on DEMA (and lack thereof on benzene) by T H F has been noted earlier in connection with the fluorescence spectra and lifetimes of these molecules. Another consideration to be made is the possibility that exciplex formation may be less exothermic between A* and B than between B* and A (see Figure 6). As implied above, the photokinetic data acquired for the DEMA/B system do not require that Scheme I contain the step describing exciplex dissociation into excited benzene and ground-state amine. This may also be a reflection of the larger endothermicity of E* B* A than dissociation into A* B. The rate constant of the latter dissociation step, symbolized by kAE,was estimated from model calculations. The parameter Y, kE kAE, is directly obtainable from the photokinetic data. Partitioning of this value into its components was attempted by using the modeling approach mentioned above. These studies > [DEMA*Io), emission characteristic of FB is observed with no evidence of sensitized DEMA fluorescence. Similarly, for A,, = 227 nm ([DEMA*Io >> [FB*],J, only DEMA fluorescence is observed; the structured FB emission is not detected. However in this system, excitation at either wavelength produces another, weak emission band at longer wavelength. For the system, [FBI = 1.6 X and [DEMA] = 4.1 X 10-3 M, excitation at 266 nm produces, in addition to FB emission at 281 nm, a broad, structureless band that extends as far as ca. 450 nm. By subtracting the maximum-normalized fluorescence spectrum of FB from the total emission, a distinct maximum is this red-shifted emission can be resolved at ca. 350 nm. Excitation at 227 nm also produces a spectrum that reveals the presence of the 350-nm feature. Photokinetic studies show that fluorescence decay, monitored at 280 nm, is single exponential for excitation at 227 nm (DEMA*) or 266 nm (FB*). In the former case [FBI was varied in the range (1-300) X IO4 M, while in the latter, [DEMA] was in the range (1-1 20) X lo4 M. In both sets of experiments, linear Stern-Volmer kinetics are observed. When monitored at longer wavelength (e.g., 380 nm), fluorescence decay is complex, corresponding to the difference in two-exponential terms. The long-lived component matches the quenched decay of the species directly photoexcited, and the short-lived component (having the negative amplitude) is concentration-independent (for [B] C 1O-* M). These observations are interpreted according to Scheme 11. The omission of "feedback" from the exciplex to either electronically excited amine or fluorobenzene is justified by (1) the
-
+
+
+
-
solvent THF
Ai
M-' s-I 0.11 0.45 0.11
0.45 0.11
'B and A denote, respectively, the benzene derivative and the amine. single-component nature of the fluorescence decay of the amine and aromatic species and (2) the linear Stern-Volmer kinetics observed. The rate constants for the DEMAJFB system in nhexane are summarized in Table I1 and will be discussed below. FluorobenzeneJDEMA in THF. The absorption and emission spectra of FB in T H F are similar to those in n-hexane. The differences in the DEMA absorption and flurescence spectra in T H F vis-8-vis n-hexane have been noted above. Excitation of a FBJDEMA mixture at 267 nm (FB*) produces only FB fluorescence; thus, unlike the situation with benzene, no sensitized DEMA fluorescence is detected. As the DEMA concentration M) the FB emission intensity is increased (1.0 X IO4 to 1.0 X decreases and another, weaker band, which is structureless, appears with increasing intensity at 400 nm. Similarly, excitation at 222 nm (DEMA*) produces DEMA fluorescence at 340 nm with no detectable FB fluorescence. As the added FB concentration is M), the DEMA fluorescence increased (1 .O X 1O-j to 2.0 X intensity decreases with a concomitant increase in a 400-nm emission band, similar in appearance to that described above. Fluorescence decay, when excited at 267 or 222 nm and monitored at 285 or 340 nm, respectively, is observed to be single exponential over the concentration range of either unexcited species used in this study. Emission, monitored at 400 nm, however, is observed to conform to the difference in two-exponential components, the short-lived component (having the negative amplitude) being concentration-independent over a limited range. In these respects, the FBJDEMA system in THF behaves similarly to that in n-hexane solution. Accordingly, the same kinetic scheme is proposed; the rate constants obtained are listed in Table 11. 1,4-DifluorobenzeneJDEMAin n-Hexane. Excitation at 267 (DFB*) or 222 nm (DEMA*) produces emission in the 280290-nm region, and because of the overlap of DFB and DEMA fluorescence spectra, definitive evidence of sensitized emission cannot be obtained. However, it is apparent that as the concentration of either one of the components is increased, fluorescence quenching occurs along with the appearance of another, broad band at ca. 370 nm. In these experiments, the concentration -5.0 X low2M and ranges of DFB and DEMA are 3.0 X 1 .O X 10d3-l.O X M, respectively. We observe similar photokinetic behavior in this system as in FBJDEMA reported above. Especially notable is that fluorescence decay, monitored a t ca. 370 nm, consists of two components, one with a lifetime corresponding to that of the directly excited species (DFB*) or (DEMA*) and the other, shorter, component having a negative amplitude. The short-lived component is concentration independent, within the sensitivity of the reconvolution analysis. The photokinetic data obtained for this system are also contained in Table 11. Conclusions In n-hexane solution, energy transfer between photoexcited DEMA and benzene occurs with diffusion control; reverse energy transfer is about 13% efficient, and no evidence of exciplex emission is revealed for moderate amine and aromatic concentrations. Although an amine/benzene exciplex is more stable than either locally excited state, as shown by the fact that exciplex emission is the dominant, perhaps exclusive, radiative decay channel in 3-phenyl-l-(dimethylamino)propane (PDMP), this exciplex is evidently kinetically unimportant in intermolecular systems where the free energy of formation appears to be too large (for moderate concentrations). Thus, in an intramolecular system
J . Phys. Chem. 1988, 92, 6625-6630 (e.g., PDMP) the exciplex acts as a trap of excitation energy. Moreover, it appears that exciplex formation demands a more stringent geometrical requirement than does collisional energy transfer. By contrast, in T H F solution, energy transfer proceeds from photoexcited benzene to the amine with no evidence of energy transfer from DEMA* to B. In this solvent, weak exciplex emission is observed in addition to fluorescence from B* or DEMA*. The fluorescence maximum of this exciplex in T H F solution in consistent with that of the intramolecular exciplex fluorescence observed in PDMP. Exciplex emission from PDMP is dominant in THF solution, as it is in n-hexane. Because of its charge-transfer character, the exciplex is stabilized in the T H F medium relative to n-hexane and is proposed to act as an intermediate in the energy-transfer process. Thus the encounter complex between B* and DEMA relaxes to a more structurally defined entity (the exciplex), which can dissociate into DEMA*
6625
and B. The formation efficiency of this exciplex is much lower between DEMA* and B than between B* and DEMA, possibly because of the diminished mobility of the solvated DEMA* vish i s DEMA. The structure of DEMA*, presumably planar, may also present steric hindrance to exciplex formation with B. The two fluorine-substituted benzenes studied, fluorobenzene and 1,4-difluorobenzene, appear to form more stable exciplexes relative to benzene, an in each case the exciplex acts as a trap of excitation energy. Exciplex emission is observed from both systems in n-hexane and THF solutions. In neither case, however, is energy transfer observed in either direction. Acknowledgment. We are grateful to the donors of the Petroleum Research Fund, administered by the American Chemical Society, for the support of this work. Registry No. DEMA, 616-39-7;FB, 462-06-6;DFB, 540-36-3;B, 71-43-2.
Fluorescence Quenching and Electron Transfer in Water-Soluble Copolymers of Methacrylic Acid and Vinylperylene or N- [10- (4-Aminonaphthalimide)]-2-met hy lacrylamide R. D. Stramel, S. E. Webber,* Department of Chemistry and Center for Polymer Research University of Texas at Austin, Austin, Texas 78712
and M. A. J. Rodgerst Center for Fast Kinetics Research, University of Texas at Austin, Austin, Texas 78712 (Received: April 6, 1988)
Copolymers of methacrylic acid and vinylperylene or N - [10-(4-aminonaphthalimide)]-2-methacrylamide (ANI) have been prepared, and the fluorescence spectroscopy of the polymers has been studied in aqueous solution. Methylviologen (MV2+) and sulfonated propylviologen (SPV) quench the fluorescence of the chromophores, resulting in charge-separated products at low pH. Yields of the reduced viologens sensitized by perylene are 0.41 and 0.36 for SPV and MV2+, respectively. For ANI these values are 0.18 and 0.07. Recombination of the charge-separated ions occurs via a second-orderprocess: kw.+IPER.+ = (2.1 0.5)X 1Olo M-ls- 1; kSPV.-/PER.+= (8.0 A 3) x 109 M-l s-1; k MV.+/ANI.+ = (9.0 0.5) X lo9 M-l s-'. k SPV.-JANI.+ = (1.1 f 0.1) X 1O'O M-I s-' (all in oxygen-free solution).
*
*
Introduction One of the most important photochemical processes is the transfer of an electron to or from an excited-state species with the formation of a highly reactive radical pair, e.g.
D* + A
-
Do+
+ A'-
(1)
The radical pair in (1) can either re-form the ground-state species or participate in other useful reactions. There have been many different systems investigated with respect to reaction 1, frequently focusing on "microheterogeneous media", such as micelles,I colloids,* membra ne^,^ and the subject of the present paper, polyelectrolyte~.~ There are two issues in reaction 1: (1) are the species D" and A'- formed at all and (2) how long do these species persist in the absence of an irreversible chemical reaction (Le., is either species a "sacrificial" electron donor or acceptor)? In the present study the existence of these species is defined by the presence of its characteristic transient absorption spectrum on the time scale of ca. 10-15 ns (our laser pulse width). W e are interested in the fraction of excited state-electron acceptor encounters that lead t Permanent address: Center for Photochemical Sciences, Bowling Green State University, Bowling Green, OH.
9
to a radical pair. The lifetime of these species is monitored by the decay of their transient absorption. The radical pair should persist a reasonably long time, on the order of tens of microseconds or longer, to carry out successive chemical reactions. We have previously reported that the singlet state of diphenylanthracene (DPA) covalently bound to poly(methacry1ic acid)5 or poly(styrenesulfonate)6 can yield a radical pair with reasonable efficiency with viologen quenchers. While the details depend on the viologen, the general conclusion was that a protonated polyacid provides "hydrophobic protection" for the aro(1) Thomas, J. K. The Chemistry of Excitation at Interfaces; ACS Monograph Series, No. 181; American Chemical Society: Washington, D.C., 1984. (2) Kalyanasundaram, K. Photochemistry in Microheterogeneous Systems; Academic: New York, 1987. (3) Fendler, J. Membrane Mimetic Chemistry; Academic: New York, 1983. (4) See forthcoming extensive review of polyelectrolytes: Rabani, J. In Photoinduced Electron Transfer; Fox, M. A,, Chanon, M., Eds.; Elsiever: in press. ( 5 ) Delaire, J. A,; Rodgers, M. A. J.; Webber, S. E. Eur. Polym. J . 1986, 22, 189. (6) Delaire, J. A,; Sanquer-Barrie, M.; Webber, S. E. JPhys. Chem. 1988, 92, 1252.
0022-3654/88/2092-6625$01.50/00 1988 American Chemical Society