2831
ISC Paths in Excited Charge-Transfer Systems
Intersystem Crossing Paths in Excited Charge-Transfer Systems N. Orbach, J. Novros, and M. Ottolenghi* Department of Phys/ca/ Chemistry, The Hebrew University. Jerusalem, lsrael
(Received May 8,7973)
Publicalion costs assisted by the U.S. National Bureau of Standards
Pulsed laser photolysis experiments are carried out in aromatic exciplex systems, investigating the generation of triplet states under varying conditions of dissolved oxygen and solvent polarity. Four different intersystem-crossing (ISC) paths, leading to the triplet states of the donor (D) or the acceptor (A), are identified: (I) intersystem crossing within the thermalized exciplex, l(A-D+)*, formed in nonpolar solvents. This is a “slow” ISC mechanism, competing with fluorescence emission from 1(A-D+)*. (11) Triplet states are generated in polar solvents uia the homogeneous recombination of solvated radical ions: A - + Ds+ + 3A* D (br 3D* A). (111) Intersystem crossing may occur in both polar and nonpolar solutions from nonrelaxed donor-acceptor pairs. This “fast” process circumvents the thermalized exciplex in hydrocarbon solvents and the separated ions in methanol or acetonitrile. (IV) Triplet states may also be formed as a result of quenching of the exciplex fluorescence by molecular oxygen: 1(A-D+)* 0 2 3A* + D + 0 2 (or 3D* + A + 0 2 ) .
+
+
+
-
With the purpose of clarifying the apparent controversy Introduction concerning the mode of triplet-state generation, we have The technique of pulsed laser photolysis has been resubmitted nine exciplex systems to a nanosecond laser cently applied to the study of charge-transfer complexes photolysis study in both polar and nonpolar solvents. Dein the excited state, leading to new information contailed light-induced spectral changes were recorded startcerning exciplexes and excited electron donor-acceptor ing -15 nsec after triggering the laser, aiming to provide (EDA) comp1exes.l A particular aspect of significant flnambiguous evidence showing that the initial spectrum, photochemical relevancy is the mechanism by which triprecorded prior to any significant decay of the exciplex lets are generated in excited, intermolecular, charge(nonpolar solvents) or the ions (polar solvents), does intransfer states. clude a triplet-state contribution. We have also studied In previous publications2 we have suggested that in the process of the exciplex fluorescence quenching by mothe cases of the exciplexes formed between N,N-diethylanlecular oxygen investigating its possible association with iline (DEA) and pyrene or anthracene, intersystem crossintersystem crossing. The results bear on the general picing occurs from nonrelaxed excited complex states. Nameture of triplet state generation during charge-transfer inly, crossing to the triplet manifold precedes the populateractions in the excited state. tion of the thermalized fluorescent exciplex, I(A-D+)*, in nonpolar solvents as well as the formation of separated Experimental Section radical ions in polar solutions. This suggestion was based The nanosecond nitrogen laser (337.1 nm) photolysis on the observation that the absorbance, at the charactertechnique6 as well as the steady-state fluorimeter3 have istic maxima of the anthracene and pyrene triplets, apbeen described previously. pears immediately after a -10-nsec laser pulse, prior to N,N-Dimethylaniline (DMA) and N,N-diethylaniline any substantial decay of the fluorescent exciplex.2 This (DEA) obtained from BDH were redistilled under nitroprefluorescence ISC mechanism, later found to be consisgen. Spectroscopic grade acetonitrile (AcN), methanol tent with temperature effects on triplet and fluorescence (both Fluka), and toluene (Matheson), as well as methylyield^,^ was also postulated by Mataga and coworkers in cyclohexane (MCH) (Fluka, puriss.), were used without the case of the tetracyanobenzene-toluene EDA complex, further purification. Pyrene, anthracene (Anth), excited within its charge-transfer absorption band.4 N,N,N’,N’,-tetramethyl-p-phenylenediamine (TMPD), Recently, the above mechanism was questioned by naphthalene (Naph), and biphenyl (Biph) were all zone Land, Richards, and tho ma^,^ who observed in the case refined. Solutions were deaerated by bubbling nitrogen or of the exciplex formed between anthracene and N,N-diby a freeze and thaw technique. All experiments were carmethylaniline (DMA) a distinct growing-in of the anthraried out at room temperature. cene triplet state, matching the decay of the exciplex fluorescence. It will be shown below that similar growing-in Results processes are also observed in the case of exciplexes involving TMPD (N,N,N’,N’-tetramethyl-p-phenylenedi- I. Deaerated Nonpolar Solutions. Pulsed laser experiamine), indicating that ISC occurs within the thermalized ments were carried out in deaerated nonpolar solvents exciplex in competition with the emission of fluorescence. such as methylcyclohexane or toluene, where the quenchThe failure to observe a triplet growing-in in other sysing of an excited donor, IA* (or acceptor, ID*), by a tems is attributed by Land, et ~ l . to, ~a coincidence of the ground-state D (or, correspondingly, A) molecule is assocorresponding triplet and exciplex absorptions, so that ciated with the formation of a fluorescent exciplex. In all transitions such as l(A-D+)* 3A* + D are not associthe presently investigated systems this quenching process ated with a net change in absorbance. was also found to be associated with intersystem crossing, s3
-
The Journalof Physical Chemistry, Vol. 77, No. 24, 7973
N. Orbach, J. Novros, and M. Ottolenghi
2832
b
'Anfhracenz, DEAin MCH
(1.
'TMPD*+ Naphthalene in MCH
c.
'Anthracene*+ DMA ip MCH
___11_11
-02
-- 0.1
380
400
420
440
A, nm
......
Figure 1. Characteristic oscillograms and transient spectra in exciplex systems prepared in deaerated nonpolar solvents by pulsed laser excitation. In each case three oscilloscope traces are reported. The upper one represents the separate contribution of fluorescence (monitoring light shutter closed) to the lower trace (absorbance fluorescence) recorded in the presence of the monitoring pulse. The lowest trace is the separate contribution of the absorption obtained by substracting the fluorescence. a, [Anth] = 10- M , [DEA] = 0.5 M . . . . X . . . . (ADO)values recorded at the end of the laser pulse (15 nsec after triggering); (AD- ) values recorded 200 nsec after triggering. (Since the triplet decay is substantial even during the first 200 nsec, we have normalized the maximum of the ADI- curve to fit that of AD0,~thusrendering easier the comparison of the two spectra); b, [TMPD] = 2 X M, [Naph] = 0.1 M; . . . X . . . . ( A D O )15 nsec; (AD - ) 350 nsec; c, [Anth] = 8 X M, (DMA] = 0.75 M; . . . X . , .. (ADo) 15 nsec (by graphical extrapolation); --e--, (AD ) 350 nsec.
+
......
4
0.3
j
0.2
Ii
- 0.1
u z
i
/.
K''"
....
R.
/ I
t---
I
"
350
1
a 0.2
0.I
400
450
'TMPD*+Biphe&l in MCH
1
1L
0.0300
X,nm Figure 2. Transient absorbance changes in the laser photolysis of exciplex systems in deaerated nonpolar solvents: - points recorded 200 nsec after triggering; . X . , , , ( A D O ) ,points recorded 50 nsec (a,b) or 10 nsec (c, by graphicai extrapolation, and d); a, [DEA] = 0.25 M, [Naph] = 0.2 M; b, [DEA] = 0.25 M, [Biph] = 0.2 M; c, [TMPD] = 2 X M, [Biph] = 0.1 M; d, [Pyrene] = M, [DEA] = 0,5 M. I
.
The Journal of Physical Chemistry, Vol. 77, No. 24, 1973
2833
ISC Paths in Excited Charge-Transfer Systems
leading to the triplet state of either donor (3D*) or acceptor (3A*). This is shown in Figures 1 and 2 where some characteristic oscillograms and flash difference spectra are presented for various donor-acceptor pairs. I t can be seen that the transient absorbance recorded a few hundred nanoseconds after the laser pulse (OD=), when most of the exciplex has decayed, consists of the characteristic triplet-triplet absorbtion bands of either A or D. When considering ODo, the "initial" change in absorbance recorded a t the end of the laser pulse (-15 nsec after triggering), one should bear in mind that a t this time most of the exciplex, whose lifetime is of the order of 80 nsec, is still present. Thus, OD0 will possibly include the contributions o f (a) the exciple^,^ (b) the relatively small amount of triplets which might have been populated bia intersystem crossing from the thermalized exciplex within the first 15 nsec, and (c) triplets populated uia fast mechanisms circumventing the thermalized exciplex. For example, we examined the lTMPD + Naph pair (Figure l b ) , where around the 415-nm peak of 3Naph* ODo < ODm, indicating a growing-in process for the naphthalene ixiplet. Outside the above spectral range OD0 > ODm,reflecting the decay of the exciplex absorbance? Within the limits of experimental accuracy the triplet growing-in rate matches the decay of the exciplex absorption or fluorescence, in agreement with the competition scheme '(A-D+)*
-+
'(A-D+)*
--+
A
+
3A*
+
D
+
D
hv
(fluorescence)
(1)
( i n t e r s y s t e m crossing) (2)
in which I(A-D+)* represents the thermalized fluorescent state of the complex. Distinct growing-in stages within the range of the triplet band were also observed (see Figures 1 and 2) for IA.nth* + DMA and TMPD* Biph. No substantial growing-in was observed for lAnth* + DEA, Naph, lDEA* Biph, and IPyrene* DEA IDEA" around the absorption maxima of the corresponding triplets. In the cases of Iperylene" DMA and lpyrene" DMA an exact growing-in analysis was prevented by the intense interfering fluorescence in the range of the triplet absorption. II. Aeareated Nonpolar Solutions Laser photolysis experiments were also carried out in aerated solutions under conditions in which 0 2 could not efficiently compete with the excited donor-acceptor interaction leading to the exciplex. Thus (see Table I) the concentrations of A or D were set so as t o assume that k(D* + A)[A] >, k(D* + 02)[02] or k(A* D)[D] >> /?(A* + 0 2 ) 0 2 ] correspondingly. In all cases we found the exciplex fluorescence to be efficientIy quenched by oxygen, the results in Table I being consistent with quenching rate constants of 1O1O I M - l sec-l. Although quenched by 0 2 , the relatively low value of their reaction rate constants ( - lo9 M-l sec-I) allowed the detection of triplet states even in aerated solutions (Figure 3). By carrying out appropriate extrapolations, the relative intersystem crossing yields were estimated as presented in Table I. Figure 3 shows two different characterNaph system, the istic oxygen effects. In the ITMPD* oxygen quenching process is associated with a drop in the triplet yields Since FIFO = DT/DT0 (where F and DT are correspondingly the fluorescence intensity and the triplet absorbancy in the presence of 0 2 , while P and DTo are the corresponding values in deaerated systems) and since we have previously shown that for lTMPD*-Naph ISC oc-
+
+
+
+
+
+
+
-
+
'Anthracene*+ DMA in MCH
' T M P d + Naphthalene in MCH
io3
x
wo 2 V z d m iT 0
:
01
d
00
380
400
420
'
X,nm
Figure 3. Oscillograms and transient spectra showing oxygen effects on triplet yields in exciplex systems. In both cases the upper oscillograms were recorded in deaerated solutions and the lower ones in air-saturated systems. Upper and lower traces are correspondingly in the absence and presence of monitoring pulse: - - 0 - -, hydrocarbon triplet absorption recorded in deaerated solutions (-300 nsec after triggering): ... X..., hydrocarbon triplet in air saturated solutions ( - 2 0 nsec after triggering). Concentrations: [TMPD] = 2 X M. [Naph] = M, [DMA] = 0.75 M . 0.2 M ; [Anth] = 8 X
curs primarily cia process 2, we may conclude that the 02-quenching reaction competes with (2) leading to deactivation according to
'(D+A-)*
+
0,
+
D
+
A
+
0,
(3)
+
By contrast, the 02-quenching effect in the IAnth* DMA system leads to an enhancement of the triplet absorbance, establishing the alternative process
YD+A-)*
+ o2
-
+-
3 ~ *
D
+
0)
(4)
Table I shows that process 3 also occurs in the case of 'TMPD* + Biph while process 4 predominates for Ipyrene* DMA, IAnth* + DMA and lAnth* -k DEA. III. Polar Solutions. Upon carrying out the CT quenching experiments in a polar solvent such as acetonitrile or methanol, no exciplex fluorescence or absorption can be detected-. Instead, the quenching process is associated with ionization, leading to the solvated radical ions A,and Ds+.8 This is confirmed for the present systems in Figures 4 and 5, showing the transient absorbance changes induced by laser excitation in polar solvents. The same figures show that, with the exception of the ITMPD* [or l(DEA)*] + Biph systems, the decay of the characteristic absorbtion bands of A,- and D,+ is associated with a matching growing-in within the range of the corresponding triplet bands, establishing the process
+
-4,-
+
D:
--+
34*
+ D (or 3D* +
A)
(5)
Such reactions have been recently investigated in chemiluminescence s t ~ d i e s . ~ The Journal of Physical Chemistry. Vol. 77, No. 24, 7973
2834
N. Orbach, J. Novros, and
'Anthracene'
C
+ DEA
M. Ottolenghi 1
in A c N
0
W
W
z a 5 0 W 0
z
a a
m 0
m
a
C
I
380
400
420
460
440
480
520
500
1
420
1
'
44c
1
1
460
,
,
480
1
5m
520
X , nm
X, nm
Figure 4. Transient spectra and characteristic oscillograms in the laser photolysis of anthracene and pyrene quenched by DEA in polar solvents. Left: M anthracene and 0.2 M DEA in acetonitrile; 15 nsec (- - X - - ) and 300 nsec ( . . , 0 . . . ) after triggering. The oscillograms represent the anthracene triplet evolution (425 nm) and the decay of the DEA+ ion (465 nm). A quantitative comparison between the triplet growing-in and the ions decay is complicated by the superimposed triplet decay which causes its growingin to look faster than the decay of the ions. Right: solid line: the transient change in absorbance recorded 25 nsec after firing the nitrogen laser in a methanol solution of M pyrene quenched by 5 X lo-' M DEA. Dashed line: the superimposed spectra of pyrene- and DEAi reproduced from available data in the literature (details in ref 3). The 415-nm band which is absent in the superimposed spectra of D+ and A- is due to the triplet state of pyrene. Insert: oscillogram taken at the maximum of the pyrene negative ion absorption band. TABLE I: Oxygen and Solvent Effects on Complex Fluorescence and Triplet Yields in Exciplex Systems Nonpolar solutions
System
'Anth* f DEA
Quencher
Triplet state obsd
Solvent
3Anth*
Toluene
r:fci
0.5
+ DEA
3Pyrene*
Toluene
0.5
'Anth" f DMA 'Pyrene* DMA 'TMPD* Naph 'TMPD* Biph IDEA* Naph 'DEA* f Biph
3Anth* 3Pyrene* 3Naph* 3Biph* 3Naph* 3Biph*
MCH MCH MCH MCH Toluene Toluene
0.75 0.5 0.2
'Pyrene*
+ +
+ +
[a] 13
Relative fluorescence T O , nsec yields, (deaeratF/Fob edIc
0.05
55
Polar solutions Relative nsec triplet ated) (aer- D Tyields / D T O ~ Solvent 00 i t;:) DT~
T,