Ultrafast intersystem crossing and spin dynamics of ohotoexcited perylene-3,4:9,10-bis(dicarboximide) covalently linked to a nitroxide radical at fixed distances

Article abstract of DOI:10.1021/ja808924f

Time-resolved transient optical absorption and EPR (TREPR) spectroscopies are used to probe the interaction of the lowest excited singlet state of perylene-3,4:9,10-bis(dicarboximide) (¹PDI) with a stable tert-butylphenylnitroxide radical (

Full text of DOI:10.1021/ja808924f

Published on Web 02/20/2009
Ultrafast Intersystem Crossing and Spin Dynamics of
Photoexcited Perylene-3,4:9,10-bis(dicarboximide) Covalently
Linked to a Nitroxide Radical at Fixed Distances
Emilie M. Giacobbe, Qixi Mi, Michael T. Colvin, Boiko Cohen,
Charusheela Ramanan, Amy M. Scott, Sina Yeganeh, Tobin J. Marks,*
Mark A. Ratner,* and Michael R. Wasielewski*
Department of Chemistry, Argonne-Northwestern Solar Energy Research (ANSER) Center, and
International Institute for Nanotechnology, Northwestern UniVersity,
EVanston, Illinois 60208-3113
Received November 13, 2008; E-mail: m-wasielewski@northwestern.edu
Abstract: Time-resolved transient optical absorption and EPR (TREPR) spectroscopies are used to probe
1*
the interaction of the lowest excited singlet state of perylene-3,4:9,10-bis(dicarboximide) ( PDI) with a
2
•
2
•
stable tert-butylphenylnitroxide radical ( BPNO ) at specific distances and orientations. The BPNO radical
is connected to the PDI with the nitroxide and imide nitrogen atoms either para (1) or meta (3) to one
another, as well as through a second intervening p-phenylene spacer (2). Transient absorption experiments
on 1-3 reveal that 1 PDI undergoes ultrafast enhanced intersystem crossing and internal conversion with
*
3*
τ = 2 ps to give structurally dependent 8-31% yields of PDI. Energy- and electron-transfer quenching of
1*
PDI by 2BPNO are excluded on energetic and spectroscopic grounds. TREPR experiments at high
•
magnetic fields (3.4 T, 94 GHz) show that the photogenerated three-spin system consists of the strongly
3*
coupled unpaired electrons confined to PDI, which are each weakly coupled to the unpaired electron on
2
•
BPNO to form excited doublet (D
1
) and quartet (Q) states, which are both spectrally resolved from the
and Q are emissive for 1 and 2 and absorptive
and Q to
. The rates of polarization transfer depend on the molecular connectivity
2
•
BPNO (D
0
) ground state. The initial spin polarizations of D
1
for 3, which evolve over time to the opposite spin polarization. The subsequent decays of D
ground-state spin polarize D
between PDI and BPNO and can be rationalized in terms of the dependence on molecular structure of
the through-bond electronic coupling between these species.
1
0
2
•
Introduction
Controlling the spin dynamics of complex multispin systems
Photoexcitation can produce well-defined initial spin states; for
example, spin-selective intersystem crossing following photo-
excitation often produces highly spin-polarized triplet states.
1
5
1
-4
is a major goal in the quest for molecule-based spintronics.
Photoexcitation of organic molecules has been shown to control
The interaction of these triplet states with stable radicals often
results in electron spin polarization of the radical, providing a
facile means for introducing and controlling spin polarization
5
-10
their properties in a wide variety of photonics applications,
while modern electron paramagnetic resonance (EPR) techniques
1
6-22
11-14
in organic materials.
A second strategy uses photoinitiated
provide an important avenue for manipulating spin systems.
ultrafast electron transfer within covalently linked organic
donor-acceptor molecules having specific donor-acceptor dis-
tances and orientations to produce highly spin-polarized radical
(
(
(
1) Rajca, A. AdV. Phys. Org. Chem. 2005, 40, 153–199.
2) Epstein, A. J. MRS Bull. 2003, 28, 492–499.
3) Awschalom, D. D.; Flatte, M. E.; Samarth, N. Sci. Am. 2002, 286,
6
7–73.
(13) Jones, J. A. In The physics of quantum information; Bouwmeester,
D., Ekert, A., Zeilinger, A., Eds.; Springer: Berlin, 2000; pp 177-
189.
(14) Schweiger, A.; Jeschke, G. Principles of pulsed electron paramagnetic
resonance; Oxford: Oxford, 2001.
(15) Levanon, H.; Norris, J. R. Chem. ReV. 1978, 78, 185–198.
(16) Kawai, A.; Shibuya, K. J. Photochem. Photobiol., C 2006, 7, 89–
103.
(
4) Wolf, S. A.; Awschalom, D. D.; Buhrman, R. A.; Daughton, J. M.;
von Molnar, S.; Roukes, M. L.; Chtchelkanova, A. Y.; Treger, D. M.
Science 2001, 294, 1488–1495.
(
(
(
5) Wasielewski, M. R. J. Org. Chem. 2006, 71, 5051–5066.
6) Verhoeven, J. W. J. Photochem. Photobiol. C 2006, 7, 40–60.
7) Segura, J. L.; Martin, N.; Guldi, D. M. Chem. Soc. ReV. 2005, 34,
3
1–47.
(
(
8) Holten, D.; Bocian, D. F.; Lindsey, J. S. Acc. Chem. Res. 2002, 35,
(17) Franco, L.; Mazzoni, M.; Corvaja, C.; Gubskaya, V. P.; Berezhnaya,
L. S.; Nuretdinov, I. A. Mol. Phys. 2006, 104, 1543–1550.
(18) Rozenshtein, V.; Berg, A.; Stavitski, E.; Levanon, H.; Franco, L.;
Corvaja, C. J. Phys. Chem. A 2005, 109, 11144–11154.
(19) Blank, A.; Levanon, H. Mol. Phys. 2002, 100, 1477–1488.
(20) Sartori, E.; Toffoletti, A.; Corvaja, C.; Garlaschelli, L. J. Phys. Chem.
A 2001, 105, 10776–10780.
5
7–69.
9) Lukas, A. S.; Wasielewski, M. R. In Molecular Switches; Feringa,
B. L., Ed.; Wiley-VCH: Weinheim, 2001; pp 1-35.
(
10) de Silva, A. P.; Guanaratne, H. Q. N.; Gunnlaugsson, T.; Huxley,
A. J. M.; McCoy, C. P.; Rademacher, J. T.; Rice, T. E. Chem. ReV.
1
997, 97, 1515–1566.
(
(
11) Mehring, M.; Mende, J. Phys.ReV. A: At., Mol., Opt. Phys 2006, 73,
52303/052301 052303/052312.
12) Harneit, W. Phys. ReV. A 2002, 65, 032322.
(21) Blank, A.; Levanon, H. J. Phys. Chem. A 2001, 105, 4799–4807.
(22) Fujisawa, J.; Ishii, K.; Ohba, Y.; Yamauchi, S.; Fuhs, M.; M o¨ bius,
K. J. Phys. Chem. A 1999, 103, 213–216.
0
3
700 9 J. AM. CHEM. SOC. 2009, 131, 3700–3712
10.1021/ja808924f CCC: $40.75  2009 American Chemical Society

Photoexcited Perylene-3,4:9,10-bis(dicarboximide)
A R T I C L E S
23-27
21,50,65
pairs (RPs) in which the initial spin state is well defined.
We
and electron spin polarization transfer (ESPT).
A triplet
and others investigated how to control the spin dynamics of these
state with a Boltzmann population of its spin sublevels can
polarize a radical by means of a spin-sorting process. Both
mechanisms are predicated on diffusive encounters between the
photoexcited triplet state molecule and the radical in solution.
In the case of the RTPM, however, it is not necessary for the
triplet state to be polarized initially. Recently, a few examples
of photoexcited triplet states having covalently attached radicals
28-35
covalent RPs using the influence of additional spins.
These
organic RPs display coherent spin motion for microseconds at room
36,37
temperature and longer at low temperatures,
which makes it
possible that this coherence could provide the basis for new organic
3,35,38-42
information-processing devices.
Spin polarization as a result of triplet-radical interactions
22,53,54,57
has been attributed for the most part to two complementary
haveappeared,e.g.,ZnTPPcoordinatedwithpyridylnitronylnitroxides
22,43-56
58,59
mechanisms, the radical-triplet pair mechanism (RTPM)
and tert-butylpyridylnitroxides,
silicon phthalocyanines and
fullerenes with attached 2,2,6,6-tetramethylpiperidine-N-oxyl
1
8,20,60-64
(
TEMPO) substituents,
as well as verdazyl and nit-
(
23) Hasharoni, K.; Levanon, H.; Greenfield, S. R.; Gosztola, D. J.; Svec,
6
5-69
ronylnitroxides attached to polycyclic aromatic molecules.
W. A.; Wasielewski, M. R. J. Am. Chem. Soc. 1995, 117, 8055–
8
056.
In many of these systems there is an ambiguity as to whether
the initial triplet state of the chromophore arises from normal
spin-orbit-induced intersystem crossing (SO-ISC) or whether
it results from a direct interaction of the radical spin with the
lowest excited singlet state of the chromophore. Quenching of
fluorescent chromophores by stable radicals has been extensively
studied in a wide variety of noncovalent and flexibly linked
(
24) Carbonera, D.; DiValentin, M.; Corvaja, C.; Agostini, G.; Giacometti,
G.; Liddell, P. A.; Kuciauskas, D.; Moore, A. L.; Moore, T. A.; Gust,
D. J. Am. Chem. Soc. 1998, 120, 4398–4405.
(
25) Weiss, E. A.; Ratner, M. A.; Wasielewski, M. R. J. Phys. Chem. A
2
003, 107, 3639–3647.
(
26) Weiss, E. A.; Ahrens, M. J.; Sinks, L. E.; Gusev, A. V.; Ratner,
M. A.; Wasielewski, M. R. J. Am. Chem. Soc. 2004, 126, 5577–
5
584.
7
0-82
(
27) Dance, Z. E. X.; Mi, Q. X.; McCamant, D. W.; Ahrens, M. J.; Ratner,
covalent systems.
However, there are very few examples
M. A.; Wasielewski, M. R. J. Phys. Chem. B 2006, 110, 25163–
of rigid systems in which the structural and electronic basis of
2
5173.
(
(
(
28) Chernick, E. T.; Mi, Q.; Vega, A. M.; Lockard, J. V.; Ratner, M. A.;
Wasielewski, M. R. J. Phys. Chem. B 2007, 111, 6728–6737.
(56) Fujisawa, J.; Ohba, Y.; Yamauchi, S. Chem. Phys. Lett. 1998, 294,
248–254.
29) Mi, Q.; Chernick, E. T.; McCamant, D. W.; Weiss, E. A.; Ratner,
M. A.; Wasielewski, M. R. J. Phys. Chem. A 2006, 110, 7323–7333.
30) Chernick, E. T.; Mi, Q.; Kelley, R. F.; Weiss, E. A.; Jones, B. A.;
Marks, T. J.; Ratner, M. A.; Wasielewski, M. R. J. Am. Chem. Soc.
(57) Fujisawa, J.; Ishii, K.; Ohba, Y.; Yamauchi, S.; Fuhs, M.; M o¨ bius,
K. J. Phys. Chem. A 1999, 103, 3138.
(58) Tarasov, V. F.; Saiful, I. S. M.; Iwasaki, Y.; Ohba, Y.; Savitsky, A.;
M o¨ bius, K.; Yamauchi, S. Appl. Magn. Reson. 2006, 30, 619–636.
(59) Tarasov, V. F.; Saiful, I. S. M.; Ohba, Y.; Takahashi, K.; Yamauchi,
S. Spectrochim. Acta, Part A 2008, 69, 1327–1330.
(60) Ishii, K.; Hirose, Y.; Fujitsuka, H.; Ito, O.; Kobayashi, N. J. Am.
Chem. Soc. 2001, 123, 702–708.
2
006, 128, 4356–4364.
31) Ishii, K.; Hirose, Y.; Kobayashi, N. J. Phys. Chem. A 1999, 103,
986–1990.
32) Mori, Y.; Sakaguchi, Y.; Hayashi, H. J. Phys. Chem. A 2000, 104,
896–4905.
33) Mori, Y.; Sakaguchi, Y.; Hayashi, H. Bull. Chem. Soc. Jpn. 2001,
(
(
(
(
(
(
(
1
4
(61) Takeuchi, S.; Ishii, K.; Kobayashi, N. J. Phys. Chem. A 2004, 108,
3276–3280.
7
4, 293–304.
34) Vlassiouk, I.; Smirnov, S.; Kutzki, O.; Wedel, M.; Montforts, F.-P.
J. Phys. Chem. B 2002, 1-6, 8657–8666.
(62) Corvaja, C.; Franco, L.; Mazzoni, M. Appl. Magn. Reson. 2001, 20,
71–83.
35) Buchachenko, A. L.; Berdinsky, V. L. Chem. ReV. 2002, 102, 603–
(63) Corvaja, C.; Maggini, M.; Prato, M.; Scorrano, G.; Venzin, M. J. Am.
Chem. Soc. 1995, 117, 8857–8858.
6
12.
36) Prisner, T.; Dobbert, O.; Dinse, K. P.; van Willigen, H. J. Am. Chem.
Soc. 1988, 110, 1622–1623.
(64) Conti, F.; Corvaja, C.; Toffoletti, A.; Mizuochi, N.; Ohba, Y.;
Yamauchi, S.; Maggini, M. J. Phys. Chem. A 2000, 104, 4962–4967.
(65) Teki, Y.; Tamekuni, H.; Takeuchi, J.; Miura, Y. Angew. Chem., Int.
Ed. 2006, 45, 4666–4670.
37) Angerhofer, A.; Toporowicz, M.; Bowman, M. K.; Norris, J. R.;
Levanon, H. J. Phys. Chem. 1988, 92, 7164–7166.
(
38) Lahti, P. M., Ed.; Magnetic Properties of Organic Materials; Dekker:
(66) Teki, Y.; Kimura, M.; Narimatsu, S.; Ohara, K.; Mukai, K. Bull.
Chem. Soc. Jpn. 2004, 77, 95–99.
New York, 1999.
(
39) Sugawara, T.; Sakurai, H.; Izuoka, A., Eds. Electronically controllable
high spin systems realized by spin-polarized donors; Gordon and
Breach: Amsterdam, 2001.
(67) Teki, Y.; Nakatsuji, M.; Miura, Y. Mol. Phys. 2002, 100, 1385–
1394.
(68) Teki, Y.; Miyamoto, S.; Nakatsuji, M.; Miura, Y. J. Am. Chem. Soc.
2001, 123, 294–305.
(
(
(
40) Miller, J. S.; Manson, J. L. Acc. Chem. Res. 2001, 34, 563–570.
41) Rajca, A. Chem. Eur. J. 2002, 8, 4834–4841.
(69) Teki, Y.; Miyamoto, S.; Iimura, K.; Nakatsuji, M.; Miura, Y. J. Am.
Chem. Soc. 2000, 122, 984–985.
42) Itkis, M. E.; Chi, X.; Cordes, A. W.; Haddon, R. C. Science 2002,
2
96, 1443–1445.
(70) Blough, N. V.; Simpson, D. J. J. Am. Chem. Soc. 1988, 110, 1915–
1917.
(
43) Bl a¨ ttler, C.; Jent, F.; Paul, H. Chem. Phys. Lett. 1990, 166, 375–
3
80.
(71) Chattopadhyay, S. K.; Das, P. K.; Hug, G. L. J. Am. Chem. Soc.
1983, 105, 6205–6210.
(44) Jenks, W. S.; Turro, N. J. Res. Chem. Intermed. 1990, 13, 237.
(45) Kawai, A.; Okutsu, T.; Obi, K. J. Phys. Chem. 1991, 95, 9130–9134.
(46) Kawai, A.; Obi, K. J. Phys. Chem. 1992, 96, 52–56.
(47) Turro, N. J.; Khudyakov, I. V.; Bossmann, S. H.; Dwyer, D. W. J.
Phys. Chem. 1993, 97, 1138–1146.
(72) Green, J. A.; Singer, L. A.; Parks, J. H. J. Chem. Phys. 1973, 58,
2690–2695.
(73) Green, S. A.; Simpson, D. J.; Zhou, G.; Ho, P. S.; Blough, N. V.
J. Am. Chem. Soc. 1990, 112, 7337–7346.
(
(
(
(
(
(
(
(
48) Goudsmit, G. H.; Paul, H.; Shushin, A. I. J. Phys. Chem. 1993, 97,
(74) Herbelin, S. E.; Blough, N. V. J. Phys. Chem. B 1998, 102, 8170–
8176.
1
3243–13249.
49) Corvaja, C.; Franco, L.; Toffoletti, A. Appl. Magn. Reson. 1994, 7,
(75) Hrdlovic, P.; Chmela, S.; Sarakha, M.; Lacoste, J. J. Photochem.
Photobiol., A 2001, 138, 95–109.
2
57–269.
50) Fujisawa, J.; Ishii, K.; Ohba, Y.; Iwaizumi, M.; Yamauchi, S. J. Phys.
Chem. 1995, 99, 17082–17084.
(76) Karpiuk, J.; Grabowski, Z. R. Chem. Phys. Lett. 1989, 160, 451–
456.
51) Hugerat, M.; van der Est, A.; Ojadi, E.; Biczok, L.; Linschitz, H.;
Levanon, H.; Stehlik, D. J. Phys. Chem. 1996, 100, 495–500.
52) Regev, A.; Galili, T.; Levanon, H. J. Phys. Chem. 1996, 100, 18502–
(77) Kollar, J.; Hrdlovic, P.; Chmela, S.; Sarakha, M.; Guyot, G. J.
Photochem. Photobiol., A 2005, 170, 151–159.
(78) Kuz’min, V. A.; Tatikolov, A. S. Chem. Phys. Lett. 1977, 51, 45–
47.
1
8510.
53) Ishii, K.; Fujisawa, J.; Ohba, Y.; Yamauchi, S. J. Am. Chem. Soc.
(79) Likhtenstein, G. I.; Ishii, K.; Nakatsuji, S. i. Photochem. Photobiol.
2007, 83, 871–881.
1
996, 118, 13079–13080.
54) Fujisawa, J.; Ishii, K.; Ohba, Y.; Yamauchi, S.; Fuhs, M.; M o¨ bius,
K. J. Phys. Chem. A 1997, 101, 5869–5876.
(80) Medvedeva, N.; Martin, V. V.; Weis, A. L.; Likhtenshten, G. I. J.
Photochem. Photobiol., A 2004, 163, 45–51.
55) Fujisawa, J.; Ohba, Y.; Yamauchi, S. J. Phys. Chem. A 1997, 101,
(81) Sartori, E.; Toffoletti, A.; Corvaja, C.; Moroder, L.; Formaggio, F.;
Toniolo, C. Chem. Phys. Lett. 2004, 385, 362–367.
4
34–439.
J. AM. CHEM. SOC. 9 VOL. 131, NO. 10, 2009 3701

A R T I C L E S
Giacobbe et al.
the quenching has been studied, and in these systems only fluore-
scence quantum yields and lifetimes have been examined.
Several photophysical mechanisms have been proposed to
account for this quenching
ing of the lowest excited singlet state of rubrene in the presence
of 4-hydroxy-TEMPO has also been attributed to EIC, which
was quite unexpected as many similar systems are quenched to
31,53,60
8
2
form large triplet yields in the presence of TEMPO.
Intermolecular quenching of chromophore excited singlet
states by EISC leads to enhanced triplet quantum yields in the
kET
8
1
*
2
•
2*PDI-• + R+
PDI + R
(1)
(2)
(3)
(4)
7
3,76-78,81
presence of a radical.
Intersystem crossing within a
chromophore is normally a spin-forbidden process but can be
kEnT
35,89
1
*
2
•
8 1PDI + 2*R•
accelerated by an organic free radical.
Again, most of the
PDI + R
work surrounding this idea is based on interactions of freely
diffusing radicals and chromophores in solution; however,
several studies of covalently linked chromophore-radical dyads
have appeared, which seek to control to some degree the
distances and orientations between the chromophore and the
radical and thereby control their electronic interactions. Green
kEISC
1
*
2
•
3*
2
•
PDI + R
8
PDI + R
kEIC
1
*
2
•
1
2 •
7
3
PDI + R
8 PDI + R
et al. studied a series of molecules in which TEMPO was
covalently bound to naphthalene by a variety of covalent
linkages and concluded that EISC via local relaxation of the
singlet, not energy transfer, is the dominant mechanism for
quenching. Intersystem crossing in the presence of TEMPO or
structurally related derivatives has also been observed for a
variety of aromatic hydrocarbons such as naphthalene and
where eq 1 is electron transfer, eq 2 is F o¨ rster and/or Dexter
energy transfer, eq 3 is electron exchange induced enhanced
intersystem crossing (EISC), and eq 4 is enhanced internal
conversion (EIC). Electron transfer from a TEMPO free radical
to the triplet state of a covalently linked 1,8:4,5-naphthalene-
bis(dicarboximide) (NI) chromophore has been reported to occur
with τcs < 15 ns in acetonitrile, resulting in a long-lived charge-
separated state with a lifetime of >200 µs. Several studies have
reported the oxidation of nitroxide radicals by fullerene triplet states,
but they are limited to intermolecular interactions.
pyrene, phenanthrene, 1,2-benzanthrene, azulene from its S
state, and perylene, as indicated by an increase in triplet quantum
2
7
2,76,78,81
83
yield.
In all of these cases, the authors note that while
EISC dominates, there may be other mechanisms contributing
to the decay of the excited singlet states.
84-86
Fluores-
Perylene-3,4:9,10-bis(dicarboximide) (PDI) and its deriva-
tives are robust organic dyes that strongly absorb visible light
and display a strong tendency to self-assemble into ordered
aggregates, thus generating significant interest as photoactive
cence quenching as a result of electron transfer to or from a stable
radical has often been considered unlikely due to the lack of a
solvent dependence, which is often key evidence for assigning this
72
mechanism. However, it has been argued that a lack of solvent
dependence is not definitive evidence for ruling out electron
transfer, since the process could be diffusion controlled and not
9
0-98
materials in a wide variety of organic electronics.
PDI
9
9
is both photochemically and thermally stable, is easily
functionalized, and serves as an excellent electron acceptor
71
appear solvent dependent. It has yet to be conclusively demon-
strated that the excited singlet state of a chromophore is quenched
by electron transfer to or from a stable radical.
1
00
with a reversible reduction potential.
Moreover, PDI has
1
01
a fluorescence quantum yield that is close to unity, so that
it has a very low intrinsic triplet yield due to SO-ISC. If a
stable radical is covalently linked to PDI at a fixed distance
and 3*PDI is formed by the interaction of PDI with the stable
radical, the dynamics of this process should be attributable
exclusively to mechanisms other than normal SO-ISC, and
Fluorescence resonance energy transfer (FRET) is generally
thought to be insignificant for nitroxide radicals as their optical
transitions in the visible spectrum have very low oscillator
1*
8
7
strengths. Yet, exceptions have been proposed, and when
energy-accepting radicals are freely diffusing in solution, it is
also necessary to consider the distribution of distances and
presence of multiple radicals in the proximity of one energy
3
*
this would provide an informative probe of how PDI
formation depends on the structure and electronic properties
of the PDI-radical system.
8
8
donor.
EIC and EISC are the two most often cited mechanisms for
excited singlet-state quenching by stable radicals. Quenching
of the singlet state of fluorescamine dyes derivatized with
nitroxides by EIC as monitored by transient absorption has been
proposed. The absence of fluorescamine triplet-triplet absorp-
tion features in the transient absorption spectrum and unfavor-
able energetics for energy transfer led the authors to conclude
that EIC is the most likely quenching mechanism. The quench-
(
89) Hoytink, G. J. Acc. Chem. Res. 1969, 2, 114–120.
(90) Tang, C. W. App. Phys. Lett. 1986, 48, 183–185.
(91) Ford, W. E.; Kamat, P. V. J. Phys. Chem. 1987, 91, 6373–6380.
74
(92) Ford, W. E.; Hiratsuka, H.; Kamat, P. V. J. Phys. Chem. 1989, 93,
6
692–6696.
(
(
(
93) W u¨ rthner, F.; Thalacker, C.; Sautter, A. AdV. Mater. 1999, 11, 754–
758.
94) Dittmer, J. J.; Marseglia, E. A.; Friend, R. H. AdV. Mater. 2000, 12,
1
270–1274.
95) Langhals, H.; Saulich, S. Chem. Eur. J. 2002, 8, 5630–5643.
(
82) Yee, W. A.; Kuzmin, V. A.; Kliger, D. S.; Hammond, G. S.;
Twarowski, A. J. J. Am. Chem. Soc. 1979, 101, 5104–5106.
83) Green, S.; Fox, M. A. J. Phys. Chem. 1995, 99, 14752–14757.
84) Araki, Y.; Luo, H.; Islam, S. D. M.; Ito, O.; Matsushita, M. M.;
Iyoda, T. J. Phys. Chem. A 2003, 107, 2815–2820.
(96) Gregg, B. A. J. Phys. Chem. B 2003, 107, 4688–4698.
(97) Neuteboom, E. E.; Meskers, S. C. J.; Van Hal, P. A.; Van Duren,
J. K. J.; Meijer, E. W.; Janssen, R. A. J.; Dupin, H.; Pourtois, G.;
Cornil, J.; Lazzaroni, R.; Bredas, J.-L.; Beljonne, D. J. Am. Chem.
Soc. 2003, 125, 8625–8638.
(
(
(
85) Borisevich, Y. E.; Kuz’min, V. A.; Renge, I. V.; Darmanyan, A. P.
IzV. Akad. Nauk SSSR, Ser. Khim. 1981, 201, 4–2019.
(98) Han, J. J.; Shaller, A. D.; Wang, W.; Li, A. D. Q. J. Am. Chem. Soc.
2008, 130, 6974–6982.
(
(
86) Samanta, A.; Kamat, P. V. Chem. Phys. Lett. 1992, 199, 635–639.
(99) Langhals, H.; Ismael, R. Eur. J. Org. Chem. 1998, 191, 5–1917.
(100) Gosztola, D.; Niemczyk, M. P.; Svec, W. A.; Lukas, A. S.;
Wasielewski, M. R. J. Phys. Chem. A 2000, 104, 6545–6551.
(101) Giaimo, J. M.; Lockard, J. V.; Sinks, L. E.; Scott, A. M.; Wilson,
T. M.; Wasielewski, M. R. J. Phys. Chem. A 2008, 112, 2322–2330.
87) Puskin, J. S.; Vistnes, A. I.; Coene, M. T. Arch. Biochem. Biophys.
1
981, 206, 164–172.
(
88) Lakowicz, J. R. Principles of Fluorescence Spectroscopy; Kluwer:
Dordrecht, 1999.
3
702 J. AM. CHEM. SOC. 9 VOL. 131, NO. 10, 2009

Photoexcited Perylene-3,4:9,10-bis(dicarboximide)
A R T I C L E S
In this paper, we present both ultrafast transient optical
absorption studies and time-resolved EPR (TREPR) studies at
high magnetic fields to probe the interaction of PDI with a stable
Samples for nanosecond transient absorption spectroscopy were
placed in a 10 mm path length quartz cuvette equipped with a
vacuum adapter and subjected to five freeze-pump-thaw degassing
cycles. The samples were excited with 6 ns, 1 mJ, 545 nm laser
pulses generated using the frequency-tripled output of a Continuum
2
•
2
•
tert-butylphenylnitroxide radical ( BPNO ). The BPNO radical
is connected via its 4-position either directly to the PDI imide
nitrogen (1) or to an intervening phenyl spacer (2). In addition,
8
000 Nd:YAG laser to pump a Continuum Panther OPO. The
excitation pulse was focused to a 5 mm diameter spot and matched
to the diameter of the probe pulse generated using a xenon flashlamp
•
the BPNO radical is attached at its 3-position directly to the
PDIimidenitrogenatom(3).Thechangesinradical-chromophore
distance and orientation within 1-3 are used to probe how the
interaction between them depends on electronic coupling.
(
EG&G Electro-Optics FX-200). The signal was detected using a
photomultiplier tube with high voltage applied to only four dynodes
(Hamamatsu R928). The total instrument response time is 7 ns and
is determined primarily by the laser pulse duration. Transient
absorption kinetics were fit to a sum of exponentials with a Gaussian
instrument function using Levenberg-Marquardt least-squares
fitting.
EPR Spectroscopy. EPR measurements at both X-band (9.5
GHz) and W-band (94 GHz) were made using a Bruker Elexsys
E680-X/W EPR spectrometer outfitted with a variable Q dielectric
resonator (ER-4118X-MD5-W1) at X-band and a cylindrical
resonator (EN-680-1021H) at W-band. For EPR measurements at
-
4
the X-band, toluene solutions of 1-3 (∼10 M) were loaded into
quartz tubes (4 mm o.d. × 2 mm i.d.), subjected to five
-4
freeze-pump-thaw degassing cycles on a vacuum line (10 mbar),
and sealed using a hydrogen torch. For EPR measurements at
-4
W-band, samples of 1-3 (∼10 M) were loaded into quartz tubes
0.84 mm o.d. × 0.6 mm i.d.) in a N -filled glovebox to a height
of ∼8 mm and sealed with a clear ridged UV doming epoxy
Epoxies, Etc., DC-7160 UV). The EPR samples were stored in a
(
2
(
freezer in the dark when not being used.
Steady-state CW EPR spectra were measured at X-band using
.2-2 mW microwave power and 0.01-0.05 mT field modulation
at 100 KHz. All TREPR measurements were made at W-band. The
samples were photoexcited at 532 nm (0.2 mJ/pulse, 7 ns, 10 Hz)
using the frequency doubled output from a Nd:YAG laser (Quanta-
Ray DCR-2) coupled to a fiber optic sample holder (Bruker E-600-
023 L). Following photoexcitation, kinetic traces of the transient
magnetization were accumulated under CW microwave irradiation
6-20 mW) with the field modulation disabled. Microwave signals
in emission (e) and/or enhanced absorption (a) were detected in
both the real and the imaginary channels (quadrature detection) and
amplified (20 MHz bandwidth, 10-90% risetime of 20 ns) to
achieve a Q/π ≈ 30 ns instrument response function (IRF), where
Q is the quality factor of the resonator and ν is the resonant
frequency. Sweeping the magnetic field gave 2D spectra versus
both time and magnetic field. For each kinetic trace, the signal
acquired prior to the laser pulse was subtracted from the data.
Kinetic traces recorded at magnetic field values off resonance were
considered background signals, whose average was subtracted from
all kinetic traces. The spectra were subsequently phased into a
Lorentzian part and a dispersive part, and the former, also known
as the imaginary magnetic susceptibility ꢀ′′, is presented.
Experimental Section
0
The syntheses and characterization of compounds 1-3 are
described in detail in the Supporting Information. All reagents were
purchased from Sigma-Aldrich and used as received. Triethylamine
2
was distilled over CaH prior to use. All final products were purified
by normal-phase preparative thin layer chromatography prior to
characterization. All solvents were spectrophotometric grade or
distilled prior to use.
Electrochemical measurements were performed on a CH Instru-
ments model 660A electrochemical workstation. All samples were
measured in a solution of 0.1 M tetra-n-butylammonium hexafluo-
1
(
6 2
rophosphate (TBAPF ) in dichloromethane purged with N to
remove oxygen. A 1.0 mm diameter platinum disk electrode,
platinum wire counter electrode, and silver wire reference electrode
were used. The ferrocene/ferrocenium couple was used as an
internal reference. Spectroelectrochemistry was performed in solu-
tions of 0.1 M TBAPF
working electrode, a platinum wire counter electrode, and a silver
wire reference electrode in a 2 mm cell under a N atmosphere.
6
in butyronitrile using a platinum mesh
2
Optical Spectroscopy. Ground-state absorption measurements
were made on a Shimadzu UV-1601 spectrophotometer. The optical
density of all samples was maintained between 0.3 and 0.6 at 532
nm (εPDI,545nm ) 46 000 M- cm )
1
-1 101
for both femtosecond and
Results
nanosecond transient absorption spectroscopy. Femtosecond tran-
sient absorption measurements were made using the 532 nm, 130
fs output from an optical parametric amplifier using techniques
Synthesis and Steady-State Characterization. The synthesis
of compounds 1-3 is summarized in Scheme 1, and the details
are given in the Supporting Information. Briefly, the synthesis
1
02
described earlier.
Samples were placed in a 2 mm path length
glass cuvette and sparged with nitrogen to prevent sample degrada-
tion. The samples were irradiated with 0.5-1.0 µJ per pulse focused
to a 200 µm spot. The total instrument response function (IRF) for
the pump-probe experiments was 180 fs. The three-dimensional
data set of ∆A vs time (0-6 ns) and wavelength (440-800 nm)
was subjected to singular value decomposition and global fitting
to obtain the kinetic time constants and their decay-associated
2
•
of BPNO -substituted PDI derivatives is carried out by
first converting N-(n-octyl)perylenemonoimide-monoanhydride
1
01
(
PIA) to N-(n-octyl)perylenediimide using urea in refluxing
1
04
DMF. Copper-promoted coupling of the resultant diimides
with either the para or the meta boronic acid of the TBDMS-
105,106
protected phenylnitroxide
followed by deprotection of the
1
03
spectra using GlobalWorks software.
(
104) Chernick, E. T.; Ahrens, M. J.; Scheidt, K. A.; Wasielewski, M. R.
(
102) Ahrens, M. J.; Sinks, L. E.; Rybtchinski, B.; Liu, W. H.; Jones, B. A.;
Giaimo, J. M.; Gusev, A. V.; Goshe, A. J.; Tiede, D. M.;
Wasielewski, M. R. J. Am. Chem. Soc. 2004, 126, 8284–8294.
103) GlobalWorks; Olis, Inc.: Bogart, GA, 2008.
J. Org. Chem. 2005, 70, 1486–1489.
(105) Baskett, M.; Lahti, P. M. Polyhedron 2005, 24, 2645–2652.
(106) Lahti, P. M.; Liao, Y.; Julier, M.; Palacio, F. Synth. Met. 2001, 122,
485–493.
(
J. AM. CHEM. SOC. 9 VOL. 131, NO. 10, 2009 3703

A R T I C L E S
Giacobbe et al.
a
Scheme 1
a
(
a) 4-Bromoaniline, pyridine, reflux, 96%; (b) tBuNO-BA, Pd(PPh
3
)
4
, Na
2
CO
3
, toluene/H
2
O, 80 °C, 12 h, 60%; (c) TBAF, THF, 0 °C, 1-2 h; (d) PbO
, NEt , O , CH Cl , reflux, 48 h.
2
,
toluene, RT, 1 h; (e) urea, DMF, reflux, 24 h, 45%; (f) tBuNO-BA(m)tBuNO-BA, Cu(OAc)
2
3
2
2
2
-
1
101
alcohol with TBAF and oxidation with lead dioxide results in
the formation of 1 and 3. The corresponding synthesis of 2 is
carried out by condensing PIA with 4-bromoaniline, followed
by a Suzuki cross-coupling of the resultant 4-bromophenyl imide
with the para boronic acid of the TBDMS-protected phenylni-
troxide. Subsequent deprotection of the alcohol with TBAF and
its oxidation with lead dioxide yields 2.
cm ), which does not change significantly in 1-3, indicating
that the electronic coupling between PDI and the nitroxide
radical is relatively weak. There is also an absorption feature
2
•
in 1-3 between 300 and 350 nm due to absorption of BPNO ,
which is accentuated by a contribution from the additional
phenyl in 2. The UV-vis spectra of PDI and p-phenyl- BPNO
alone are shown in Figure S1 (Supporting Information). BPNO
2
•
•
2
Figure 1 shows the UV-vis absorption spectra for molecules
-1
1
-3. PDI has a strong absorption at 545 nm (ꢀ ) 46 000 M
(107) Weller, A. Z. Phys. Chem. 1982, 133, 93–98.
3
704 J. AM. CHEM. SOC. 9 VOL. 131, NO. 10, 2009

Photoexcited Perylene-3,4:9,10-bis(dicarboximide)
A R T I C L E S
4
50 and 775 nm and 0 and 6 ns in both toluene and THF were
subjected to singular value decomposition and global fitting to
obtain the principal kinetic components and their decay associ-
ated spectra, Figures 2 and S3 (Supporting Information). Each
transient data set can be fit to two exponential decay compo-
nents, a major component associated with the transient spectrum
1*
of PDI including its prominent stimulated emission feature
near 615 nm and a second minor component that lives .6 ns.
The decay-associated spectrum of this long-lived species is
consistent with the known spectra and extinction coefficients
3*
91
of PDI. Nanosecond transient absorption spectroscopy was
used to confirm this assignment (Figures 3 and S4 (Supporting
3
*
Information)). For 1-3 in both toluene and THF, the PDI
transient spectral features appear at 480 nm with time constants
Figure 1. UV-vis spectra of compounds 1-3 in toluene solution.
Table 1. Redox Potentials and Free Energies of Charge
1*
R
(τ ) that are very similar to those of the ultrafast PDI decay
3*
component (τD1), while PDI decays (τD2) on the order of
2
•
1*
Separation and Recombination for the Reaction BPNO + aPDI
+
•-
microseconds (Table 2).
Using the decay-associated spectra, triplet yields (Table 2)
f BPNO + PDI (V vs saturated calomel electrode (SCE))
-
∆GCS
-∆GCR
3*
were calculated from the ratio ∆A480nm( PDI)·ε505nm/∆A505nm
-
a
ox
a
red1
a
red2
molecule
E
E
E
toluene
THF
toluene
THF
1*
(
PDI)·(ε′480nm - ε480nm), where ε505nm and ε480nm are the
PDI
BPNO
-0.61 -0.85
0.82 -1.25
extinction coefficients of the PDI ground state at 505 nm (3.2
2
•
4
-1
-1
4
-1
-1 101
×
10 M cm ) and 480 nm (1.5 × 10 M cm ), while
1
2
3
0.93 -0.66 -0.86
0.83 -0.72 -0.93
0.00 -0.50
0.11 -0.49
2.23
2.34
2.18
1.73
1.74
1.72
3*
ε′480nm is the extinction coefficient of PDI at 480 nm (4.0 ×
4
-1
-1 91
0.95 -0.68 -0.89 -0.05 -0.51
10 M cm ).
EPR Spectroscopy. The steady-state EPR spectra of 1-3
consist of a dominant triplet centered at g ) 2.0057 (Figure 4).
This triplet is due to the large hyperfine splitting of the nitroxide
N nucleus, while the smaller splittings are due to the protons
on the ring attached to it. The hyperfine splittings for 1-3 are
listed in Table 3 and were obtained from simulations of the
a
Measured in CH Cl with 0.1
2
2
M
tetra-n-butylammonium
hexafluorophosphate with a ferrocene internal reference.
14
has a very weak absorption with a maximum at 475 nm that is
typical of the n,π* transition of nitroxide radicals. The PDI
chromophore is normally highly fluorescent (see Figure S1
7
2
1
01
1
10
(
Supporting Information), Φ ) 0.98); however, its fluores-
F
spectra using WINSIM.
Figure 4) are due to very small hyperfine splittings (0.02 mT)
resulting from protons on the second phenyl group that are ortho
Additional lines observed for 2
cence is almost completely quenched in 1-3.
(
The one-electron redox potentials of 1-3 were obtained using
2
•
cyclic voltammetry (Table 1). BPNO itself undergoes revers-
ible oxidation and reduction at 0.82 and -1.25 V vs SCE,
respectively. Spectroelectrochemistry shows that one-electron
2
•
to the point at which BPNO is attached. In all three molecules
spin leakage into the PDI chromophore is small, so that no
hyperfine splittings due to the nitrogens and protons within PDI
are observed.
•
-
reduction of 1-3 produces PDI (shown for 2 in Figure S2
Supporting Information)), which displays a sharp feature at 720
(
TREPR spectra of 1 at W-band, shown in Figure 4, reveal
an initial strongly emissive feature (Q) at 3369.3 mT (g )
nm, while one-electron oxidation shows no distinct absorption
features between 400-800 nm associated with the oxidation
of the radical. The free energies of the hypothetical charge
separation and recombination reactions involving oxidation of
BPNO by PDI, calculated using the Weller formalism
Supporting Information), are also given in Table 1.
2
.0040) and a weaker emissive feature (D
1
) at 3372.0 mT (g )
2.0025), which appear within the 30 ns IRF of the spectrometer
following a 532 nm, 7 ns laser pulse at room temperature. A
third emissive signal (D ) with a hyperfine splitting of a )
2
•
1*
107
0
N
(
2
•
1
.2 mT characteristic of the nitrogen in BPNO appears at
366.5 mT (g ) 2.0057) with a somewhat slower time constant
) 120 ( 20 ns. At 260 ns, the Q signal is also split into a
triplet with a ) 0.4 mT. Each signal evolves in time from
emissive to absorptive polarization, which ultimately decays
Figure S5 (Supporting Information) and Table 4). Similar
behavior is observed following photoexcitation of 2, in which
the signals labeled Q and D exhibit emissive polarization
and evolve to absorptive polarization at longer times, while
Transient Absorption Spectroscopy. The ultrafast excited-
3
state dynamics of molecules 1-3 in both toluene and THF were
studied using femtosecond transient absorption spectroscopy.
The transient spectra are summarized in Figures 2 (toluene) and
S3 (THF; Supporting Information). Photoexcitation of the PDI
chromophore in 1-3 with a 532 nm, 130 fs laser pulse results
in bleaching of the ground-state absorption band of PDI at 545
nm that occurs within the instrument response function (IRF)
of the laser, as well as the appearance of a stimulated emission
feature near 615 nm, an absorption band with a maximum near
00 nm, and broad absorption features between 450 and 775
nm that persist >6 ns. It is well known that PDI, PDI, and
PDI all have absorption features in the 700-750 nm wave-
length region.
τ
1
N
(
1
2
•
the signal attributed to BPNO remains emissive over the
course of the experiment. The kinetics of all three signals
are slower than those observed for 1. The same three signals
appear in the TREPR W-band data for 3 except that the initial
polarization of all three signals is absorptive and evolves into
emissive polarization, which is opposite to those for 1 and
2. The kinetics for all three signals in 3 are similar in
7
1*
3*
•-
100,101,108,109
The transient spectra for 1-3 between
(
108) Rybtchinski, B.; Sinks, L. E.; Wasielewski, M. R. J. Phys. Chem. A
004, 108, 7497–7505.
109) Rachford, A. A.; Goeb, S.; Castellano, F. N. J. Am. Chem. Soc. 2008,
30, 2766–2767.
2
(
1
(110) Duling, D. R. J. Magn. Reson. B 1994, 104, 105–110.
J. AM. CHEM. SOC. 9 VOL. 131, NO. 10, 2009 3705

A R T I C L E S
Giacobbe et al.
Figure 2. (Left) Transient absorption spectra of compounds 1-3 in toluene solution at the indicated times following a 130 fs, 532 nm laser pulse. (Right)
Decay-associated spectra and their respective time constants for compounds 1-3 obtained by global analysis of the transient absorption data.
Table 2. Transient Absorption Species Associated Kinetics for
-3 at 295 K
1-4: electron transfer, energy transfer, EISC, or EIC. We
1
2
•
modified the PDI- BPNO distance, orientation (para vs meta),
and solvent polarity for 1-3 to aid in determining the relative
contributions of these mechanisms to the overall quenching of
1*
3*
3*
compound
τ
D1 (ps), PDI
τ
R
(ps), PDI
τ
D2 (µs), PDI
φ
Τ
1
(TOL)
(THF)
(TOL)
(THF)
(TOL)
(THF)
2.0 ( 0.2
2.3 ( 0.2
1.9 ( 0.2
2.1 ( 0.2
2.2 ( 0.3
2.4 ( 0.3
2.0 ( 0.2
2.4 ( 0.2
2.0 ( 0.2
2.2 ( 0.2
2.7 ( 0.3
3.0 ( 0.3
0.54 ( 0.01
0.35 ( 0.01
8.4 ( 0.5
0.24 ( 0.02
0.23 ( 0.02
0.12 ( 0.02
0.08 ( 0.02
0.29 ( 0.02
0.31 ( 0.02
1*
1
2
2
3
3
PDI. Changing the distance and orientation of the nitroxide
group relative to its point of attachment to PDI changes the
electronic coupling between these two species, while varying
the solvent polarity probes the role of electron transfer as a
quenching mechanism.
4.5 ( 0.5
0.80 ( 0.01
0.44 ( 0.01
The data in Table 1 show that electron transfer from BPNO•
2
magnitude to those for 1 with the exception of τ
3
for D
0
,
to 1*PDI in 1-3 is energetically unfavorable in toluene (∆G CS
which is not observed (Table 4).
=
0 eV), whereas it is strongly favored in THF (∆GCS = -0.5
eV). These free energies of reaction predict that there should
Discussion
be a strong solvent dependence in the observed transient kinetics
1*
for both PDI decay at 615 nm and for the formation and decay
Ultrafast Excited-State Dynamics. The femtosecond transient
absorption spectra and kinetics of 1-3 show that PDI decays
very rapidly in the presence of BPNO compared to its intrinsic
1*
•-
of PDI at 700 nm, if electron transfer occurs with a significant
2
•
quantum yield. The kinetic data in Table 2 reveal no significant
changes between toluene and THF. Given that the charge
separation reaction should be in the normal region of the Marcus
1
01
4
.5 ns lifetime. As noted above, this rapid decay could be a
consequence of any or all of the mechanisms outlined in eqs
3
706 J. AM. CHEM. SOC. 9 VOL. 131, NO. 10, 2009

Photoexcited Perylene-3,4:9,10-bis(dicarboximide)
A R T I C L E S
Figure 3. Transient absorption spectra of compound 1 in toluene and THF solutions at the indicated times following a 7 ns, 545 nm laser pulse.
Figure 4. CW EPR spectra at X-band and W-band TREPR spectra of compounds 1-3 in toluene solution following a 7 ns, 532 nm laser pulse. All spectra
are obtained at 295 K.
1
11
2
•
rate vs free energy curve, a relatively large solvent effect is
expected. Similarly, charge recombination should occur in the
Marcus inverted region, which implies that the charge recom-
bination reaction rates as indicated by the kinetics at 700 nm
should be substantially faster in THF. Neither kinetic component
observed at 700 nm shows this behavior, indicating that charge
separation and recombination do not occur competitively with
the other photophysical processes in 1-3.
absorption of BPNO overlaps poorly with the PDI fluorescence
emission band. The PDI fluorescence slightly red shifts and
becomes broader in more polar solvents; however, the absor-
2
•
bance of BPNO is not affected by changes in solvent polarity,
so that the overlap diminishes. The FRET time constants and
efficiencies for 1-3 can be estimated from the BPNO -PDI
distance, r, and eqs S2 and S3 (Supporting Information) (Table
5). The calculated FRET time constants are much slower than
2
•
1*
2
•
1*
Energy transfer from PDI to BPNO is another possible
quenching mechanism. The rate constant for FRET by the
F o¨ rster dipole-dipole mechanism depends on the spectral
overlap between the emission of the donor and the absorption
the experimental time constants for PDI decay (Table 2), so
that it is unlikely that F o¨ rster energy transfer is responsible for
the observed ultrafast quenching of 1*PDI in 1-3.
Considering Dexter-type electron exchange driven energy
transfer, the rate of this process does not explicitly depend
on the absorptivity of the acceptor, but it does depend on spectral
8
8
112
of the acceptor (eq S1; Supporting Information). Figure S1
Supporting Information) shows that the very weak n,π
*
(
(
111) Marcus, R. A. J. Chem. Phys. 1956, 24, 966–978.
(112) Dexter, D. L. J. Chem. Phys. 1953, 21, 836–850.
J. AM. CHEM. SOC. 9 VOL. 131, NO. 10, 2009 3707

A R T I C L E S
Giacobbe et al.
Table 3. EPR Hyperfine Couplings (mT)
a
Additional splittings (0.02 mT, 2H) are observed.
Table 4. TREPR Kinetics in Toluene at 295 K
Q (g ) 2.0040)
D
1
(g ) 2.0025)
D
0
(g ) 2.0057)
(ns)
compound
τ
1
(ns)
τ
2
(ns)
τ
1
(ns)
τ
2
(ns)
τ
1
(ns)
τ
2
3
τ (ns)
1
2
3
140 ( 10
210 ( 10
150 ( 10
410 ( 10
>3000
740 ( 30
46 ( 10
200 ( 10
32 ( 8
620 ( 90
>3000
800 ( 400
120 ( 20
200 ( 60
110 ( 10
170 ( 30
>3000
350 ( 30
>2000
2
6,91,109,113,114
Table 5. Estimated F o¨ rster FRET Time Constants and
complexes attached to PDI.
In addition, a variety
a
Efficiencies
of electron donor-acceptor systems in which PDI acts as an
electron acceptor have allowed access to 3*PDI by radical pair
intersystem crossing, followed by charge recombination of the
triplet radical pair. The yields of 3*PDI vary widely in these
systems and depend on the yields of the precursor radical pair
toluene
THF
compound
r (Å)
τ
EnT (ps)
φ
EnT (%)
τEnT (ps)
φ
EnT (%)
1
2
3
11.4
15.7
10.0
54
360
32
99
93
99
67
455
40
99
91
99
2
6,113,114
states.
The ultrafast formation of 3*PDI from 1*PDI is most likely
a
Calculated using PhotochemCAD 2.1 as described in the Supporting
89
2
•
due to the EISC mechanism. In this mechanism BPNO mixes
with 1*PDI to produce an excited state with an overall doublet
Information.
2
•
3*
multiplicity (D
2
); similarly, BPNO mixes with PDI resulting
) and quartet (Q) multiplicities,
in states having both doublet (D
1
overlap, so that the poor overlap between the PDI emission and
Figure 5. Note that the three-spin D and Q states both have
two unpaired electrons on PDI, so that their optical transient
absorption spectra are indistinguishable from that of PDI. The
electron exchange interaction between PDI and BPNO serves
as the first-order perturbation that drives EISC. The magnitude
of this perturbation, and thus the overall intersystem crossing
rate, depends strongly on the electronic overlap between the
orbitals that contribute to PDI and the singly occupied
molecular orbital (SOMO) of BPNO . Increasing the distance
1
2
•
BPNO absorption spectra discussed above necessarily make
3*
Dexter energy transfer slow for 1-3 as well. Moreover, energy
1*
2
•
transfer requires that the energy of the excited doublet state
2
*
•
1*
BPNO (∼2.4 eV) is below that of PDI (2.23 eV), which is
not the case. Thus, the requirements for ultrafast, efficient energy
transfer from 1 PDI to BPNO are not met for 1-3, so that it
*
2
•
1*
is unlikely this mechanism can account for the ultrafast decay
2
•
1*
of PDI. In fact, only one conclusive example of energy transfer
from an emitting chromophore to a nitroxide radical has been
2
•
1*
between BPNO and PDI should diminish this overlap and
result in an overall reduction in the yield of 3 PDI (i.e., D
*
and
8
7
1
previously reported, which is likely due to the very low
7
2
Q). Comparing the data for 1 and 2 in toluene, a 4.3 Å increase
in the distance between the radical and the chromophore results
in about a factor of 2 decrease in 3*PDI yield. In molecule 1,
extinction coefficients of most nitroxide radicals in the visible.
The positive transient absorption changes observed near 700
1*
3*
•-
2
•
115
nm may have contributions from PDI, PDI, and/or PDI ,
the spin density at the 4-position of BPNO is about 0.1,
3*
91
while those observed near 480 nm are due largely to PDI.
which further delocalizes into the adjacent phenyl ring in 2.
This places enough π spin density within the phenyl ring
attached to PDI to produce sufficient electronic coupling
between the radical and 1*PDI to drive ultrafast EISC. Placing
1*
The decay of PDI is accompanied by ultrafast formation of
3
*
PDI with yields ranging from 0.08 to 0.31, as indicated by
3*
the formation of the PDI transient spectrum with time constants
probed at 480 nm comparable to those for the decay-associated
spectrum of 1*PDI. Normally, the yield of PDI produced by
SO-ISC is <0.01, as indicated by the near unity quantum yield
of PDI fluorescence. Higher yields of PDI have been observed
using triplet sensitizers and heavy atom effects of metal
(
(
(
113) Ahrens, M. J.; Kelley, R. F.; Dance, Z. E. X.; Wasielewski, M. R.
Phys. Chem. Chem. Phys. 2007, 9, 1469–1478.
3*
114) Prodi, A.; Chiorboli, C.; Scandola, F.; Iengo, E.; Alessio, E.; Dobrawa,
R.; W u¨ rthner, F. J. Am. Chem. Soc. 2005, 127, 1454–1462.
115) Shultz, D. A.; Gwaltney, K. P.; Lee, H. J. Org. Chem. 1998, 63,
769–774.
3*
3
708 J. AM. CHEM. SOC. 9 VOL. 131, NO. 10, 2009

Photoexcited Perylene-3,4:9,10-bis(dicarboximide)
A R T I C L E S
state. This process is due to enhanced vibrational relaxation as
7
2,119
has been discussed earlier.
Time-Resolved EPR. Photoexcitation of the formally doublet
ground-state D produces the excited-state D in which the two
0
2
electrons confined to PDI are spin paired (Figure 5). Our
transient optical absorption data show that EISC occurs in ∼2
ps in 1-3, which is much faster than typical SO-ISC (e.g., the
rate for 1*PDI f PDI is only ∼10 s ), so that this ultrafast
3*
6
-1
process is most likely driven by the exchange interactions
between each of the two spin-paired electrons in PDI and the
1*
2
•
89
unpaired electron on BPNO . The resulting three-spin system
is best described at high magnetic fields by mixing the T+1, T
0
,
and T-1 eigenstates of 3*PDI with the R and ꢁ spin states of
2
•
BPNO to yield excited doublet (D
Figure 5). Transitions from the spin sublevels of D
(Figure 5, thick red arrow) are partially allowed because
the initial and final state have the same overall multiplicity,
whereas the transitions from the spin sublevels of D to Q
Figure 5, thin red arrow) involve a change in multiplicity,
so that they are formally forbidden. Therefore, the D f D
transition is expected to be more rapid than the D f Q
1
) and quartet (Q) states
(
2
to those
of D
1
Figure 5. Energy levels created after doublet-triplet mixing. D
2
f D
1
2
2
internal conversion is more rapid (thick red arrow) than D f Q intersystem
(
crossing (thin red arrow). The levels shown are for 3J > 0.
2
1
2
2
•
BPNO in a meta orientation relative to PDI, as in 3, decreases
transition, resulting in a larger initial population of D relative
1
2
•
the π spin density at the carbon atom at which BPNO is
attached to PDI, thereby reducing the overall π-electron coupling
between BPNO and PDI. However, the measured quantum
to that of Q. Following the formation of D and Q, the applied
1
microwave field induces transitions between the ∆m ) (1
S
2
•
1*
levels within the D , Q, and D manifolds.
1
0
3*
22,43-56
yields of PDI for 3 in both toluene and THF solution are about
5% higher than those for 1, so that other pathways to enhance
The conventional RTPM
spin polarization mechanism
2
derived from the three-spin interaction of an excited-state
chromophore and a radical requires that the two species interact
in a diffusive encounter. The chromophore-radical distances
and the electronic coupling between the chromophore and the
radical in 1-3 are controlled using rigid covalent linkages, so
that the RTPM mechanism of CIDEP is precluded. The
alternative ESPT mechanism requires the polarized triplet state
of the chromophore to transfer its polarization to the radical
2
•
1*
the electronic coupling between BPNO and PDI must be
operative. In addition to the through-bond π interaction, the spin
of the unpaired electron on the oxygen atom of the nitroxide
can also delocalize through the σ-bond framework by a spin
polarization mechanism.
pathway between BPNO and PDI in 3 relative to 1 may
enhance the coupling between them.
1
16,117
•
As a consequence, the shorter σ
2
1*
1
20
Given that EISC is driven by a first-order perturbation, the
and is thought to occur through either direct spin exchange
energy gap between 1 PDI and PDI is also important. Since
the energy gap between these two states is about 1.2 eV, one
might expect that EISC would be slower than observed.
*
3*
121
or energy transfer. Direct spin exchange is ruled out because
it requires the initial polarization of BPNO to be the same in
9
1
2
•
both 1 and 3, whereas the observed initial polarization of
1
18
2
•
However, semiempirical ZINDO/S calculations place the T
2
BPNO in 1 is opposite to that in 3. Energy transfer is ruled
state of PDI (3 PDI) nearly isoenergetic with PDI. This makes
**
1*
out for reasons already given.
The TREPR spectra of 1-3 at W-band all show a spin-
polarized signal centered at g ) 2.0057 (Figure 4), which is
1*
3**
it more likely that EISC occurs between PDI and PDI,
followed by very rapid internal conversion from PDI to PDI.
The states resulting from mixing PDI and BPNO are not
shown in Figure 5, but their initial spin configurations would
be carried over to D
3**
3*
3**
2
•
14
split into three lines having the same hyperfine coupling to
N
2
•
as does the ground-state signal of BPNO and is thus assigned
to that radical. If the exchange interaction is much larger than
the Zeeman interaction that splits the spin sublevels (|3J| .
, Figure 5), the D and Q states can be adequately described
0 1
The measured exchange
interaction between two BPNO radicals appended to the same
1
and Q shown in the figure because of the
ultrafast nature of this internal conversion process.
3*
The PDI lifetimes of 1-3 observed by nanosecond transient
absorption spectroscopy are all substantially shorter than the
intrinsic 3*PDI lifetime of 100 µs and display only small
solvent effects. The nanosecond transient spectra show no
evidence for electron transfer from the triplet state to yield
gꢁB
1
22
by separate spin Hamiltonians.
91
2
•
-1
123
carbon atom of an ethylene spacer is J ) -24 cm , whereas
replacing the ethylene with a meta-phenyl spacer yields J )
+
-•
124
BPNO -PDI , which is reasonable given that the energy of
-1
3*
2
•
-
4 cm . The PDI- BPNO distances and number of bonds
3*
91
PDI is only 1.2 eV, so that even in polar solvents ∆GCS
. The 2-fold change in PDI lifetime as a function of solvent
>
2 •
joining them in 1-3 are comparable to those in the BPNO
biradical with the ethylene spacer, so that we expect that J
3*
0
is most likely a result of a small change in nonradiative rate of
3*
3*
PDI. The PDI lifetimes of 1-3 will be discussed further in
the context of the TREPR kinetics presented below.
(
119) Braun, A. M.; Hammond, W. B.; Cassidy, H. G. J. Am. Chem. Soc.
1
969, 91, 6196–6197.
3*
Since the PDI yields for 1-3 are all e0.31, ultrafast electron
(120) Obi, K.; Imamura, T. ReV. Chem. Interm. 1986, 7, 225–242.
(121) Jenks, W. S.; Turro, N. J. Res. Chem. Intermed. 1990, 13, 237–300.
1*
exchange driven EIC also rapidly returns PDI to the ground
(
122) Bencini, A.; Gatteschi, D. Electron paramagnetic resonance of
exchange coupled systems; Springer-Verlag: Berlin, New York, 1990.
(
116) McConnell, H. M.; Chesnut, D. B. J. Chem. Phys. 1958, 28, 107–
(123) Shultz, D. A.; Boal, A. K.; Lee, H.; Farmer, G. T. J. Org. Chem.
1999, 64, 4386–4396.
1
17.
(
(
117) Colpa, J. P.; de Boer, E. Mol. Phys. 1964, 7, 333–348.
118) Hyperchem; Hypercube Inc.: Gainesville, Florida.
(124) Kanno, F.; Inoue, K.; Koga, N.; Iwamura, H. J. Phys. Chem. 1993,
97, 13267–13272.
J. AM. CHEM. SOC. 9 VOL. 131, NO. 10, 2009 3709

A R T I C L E S
Giacobbe et al.
between the electrons on 3*PDI and BPNO should also be
2
•
transitions. Thus, intersystem crossing results in greater depletion
-1
comparable, thus making |3J| . gꢁB
0
(∼3 cm at the W-band).
of the |D
1 1
(-1/2)〉 population relative to that of |D (1/2)〉 as
Under this condition both the g factors and the hyperfine
indicated in Figure 6a. The remaining relative transition rates
are also governed by eq 10, so that the relative sublevel
couplings exhibited by spin sublevel transitions within D
1
and
1
22
Q can be approximated by
populations of D
30 ns, as evidenced by the observed initial net emissive
polarization of both the D and Q signals. The emissive character
of the signals from both D and Q within 1 and 2 strongly
1
and Q are established prior to our IRF of
∼
1
3
4
gD1 ) - g + gT
(5)
(6)
(7)
(8)
1
R
3
1
support ferromagnetic ordering for these spin manifolds (3J >
0), as illustrated in Figures 5 and 6. In turn, the initially all
absorptive signals observed for 3 are consistent with antiferro-
magnetic ordering (3J < 0). The change in the sign of the
exchange interaction in going from a para to a meta substitution
pattern in a benzene ring is consistent with spin polarization of
N
D1
1
3
N
R
a
) - a
1
3
2
3
gQ ) gR + gT
1
3
N
R
electron density at the carbon atoms within the HOMOs and
N
Q
a )
a
127,128
LUMOs of substituted benzenes.
The TREPR spectra of
these transitions at 295 K in toluene do not exhibit their full
anisotropy because the molecules are tumbling faster than the
time scale corresponding to the energy differences between the
spin sublevels.
Using eqs 5 and 7 and the g factor of 3*PDI (g ) 2.0033,
measured independently from electron donor-acceptor systems
in which 3*PDI is produced by charge recombination of spin-
1
25,126
correlated radical pairs
and Q states correspond very well to the observed signals at g
2.0025 and 2.0040, respectively. The three transitions
), the predicted g factors of the D
1
-1
The ∼200 cm thermal energy at 295 K most likely exceeds
|
3J| significantly, so that reverse intersystem crossing from Q
)
to D
ground-state D
1
is fully competitive with the decay of these states to
and relaxation of their spin populations back
expected from Q overlap strongly because of the small degree
0
of anisotropy observed in the spectra. However, the observed
to a Boltzmann distribution. This reverse intersystem crossing
has been termed the reverse quartet mechanism (RQM) and used
N
signal for Q in 1-3 is split into a triplet for which a
Q
) 0.4
mT as predicted by eq 8. Given the limitations of signal-to-
noise and spectral line width, aN
is not apparent at all times. In
addition, the spectral line due to D is lifetime broadened in
is not observed. The ability of this approximate
to describe the TREPR spectra and kinetics of a tetramethylpi-
18
Q
peridinyl-N-oxyl (TEMPO) radical covalently bound to C60
.
1
For molecules 1 and 2, when 3J > 0, the initial spin level
N
1-3, so that a
D
1
populations of D and Q (Figure 6a) evolve in time as the
1
treatment to predict the g factors and hyperfine couplings of
the observed D and Q signals supports the assumption that |3J|
gꢁB
For 1 and 2, the TREPR spectra at early times show that all
three signals due to D , D , and Q are initially polarized in
emission, while for 3 all three signals are polarized in absorption.
Even though D and Q are well separated by 3J, the zero-field
splitting interaction, HZFS, of the two unpaired electrons localized
transitions from the spin sublevels of Q back to D
b). This redistribution of spin population within D
coupled with the relative rates of depletion of the D
populations to ground-state D
TREPR transitions at later times. Focusing again on the two
sublevels that are closest in energy (Figure 6b), for |Q (+3/2)〉
1
occur (Figure
1
6
1
and Q
and Q
.
0
.
1
0
results in the observed absorptive
0
1
1
f |D
)〉 that was derived from both |D
returned preferentially to |D (-1/2)〉, once again due to the
smaller gap between |D (-1/2)〉 and |Q(+3/2)〉. A similar
1
(-1/2)〉 the relatively large initial population in |Q(+3/
2
1
(-1/2)〉 and |D (+1/2)〉 is
1
on PDI mixes D
1
and Q
1
2
z
2
2
x
2
y
2
z
2
1
HZFS ) D (S - S /3) + E (S - S ) ) D (S - S /3) +
qualitative analysis can be made for the remaining spin
sublevels. The same mechanistic picture prevails for 3, except
that the spin manifold of Q is now energetically above that of
D (i.e., 3J < 0), and the two states closest in energy are |D (+1/
2
2
E (S + S )/2 (9)
+
-
where D and E are the zero-field splitting parameters. Thus,
1
1
the rate of intersystem crossing from the spin sublevels of D
to those of Q is given by
1
2)〉 and |Q(-3/2)〉.
The kinetics for the formation and decay of the spin-polarized
TREPR signals of D , Q, and D are a complex mix of
1
0
2
|
〈Q(m )|H |D (m )〉|
S ZFS 1 S
population changes between the spin sublevels and spin
relaxation processes. This problem was initially treated using a
modified Bloch equation approach including kinetic terms
k(|D (m )〉 f |Q(m )〉) ) k ·FC·
1
s
s
0
2
(E (m ) - E (m ))
Q S D S
1
(
10)
connecting the spin sublevels of D
ground-state D .
0
1
and Q to those of the
This model yields a complex series of
equations having many rate constants that are not readily
1
8
0
where k is the fully spin-allowed, vibrationally limited rate
13
-1
constant (∼10 s ), FC is the Franck-Condon factor for the
5
8,59
accessible experimentally. Tarasov et al.
recently applied
transition, and the final term is the spin-dependent transition
1
strength resulting from the HZFS perturbation that mixes D and
Q. The Franck-Condon terms should be similar (and large)
for all transitions because the energy gaps between the spin
sublevels are all much smaller than those between typical
vibronic transitions.
(125) Dance, Z. E. X.; Mickley, S. M.; Wilson, T. M.; Ricks, A. B.; Scott,
A. M.; Ratner, M. A.; Wasielewski, M. R. J. Phys. Chem. A 2008,
1
12, 4194–4201.
(
(
126) Zeidan, T. A.; Carmieli, R.; Kelley, R. F.; Wilson, T. M.; Lewis,
F. D.; Wasielewski, M. R. J. Am. Chem. Soc. 2008, 130, 13945–
13955.
Referring to Figure 6a, the intersystem crossing rate for 1
127) Thompson, A. L.; Ahn, T.-S.; Thomas, K. R. J.; Thayumanavan, S.;
Martinez, T. J.; Bardeen, C. J. J. Am. Chem. Soc. 2005, 127, 16348–
16349.
and 2 is greatest for |D
that the energy gap between these states (E
)) is the smallest for any of the six allowed (∆m
1
(-1/2)〉 f |Q (+3/2)〉 due to the fact
(+3/2) - ED1(-1/
) (1, (2)
Q
2
S
(128) Zimmerman, H. E. J. Phys. Chem. A 1998, 102, 5616–5621.
3
710 J. AM. CHEM. SOC. 9 VOL. 131, NO. 10, 2009

Photoexcited Perylene-3,4:9,10-bis(dicarboximide)
A R T I C L E S
Figure 6. D
times. The levels shown are for 3J > 0.
1
-Q intersystem crossing resulting in emissive polarization from D
1
and Q at (A) early times followed by enhanced absorption at (B) later
the stochastic Liouville equation to this problem taking into
account the relevant kinetic terms in the relaxation operator and
considering the full anisotropy of the Zeeman and dipolar
interactions. This elegant approach accounts for the case in
which the three-spin system may be tumbling in solution at a
rate that is comparable to the transition rates between the spin
manifolds as well as the spin relaxation rates. This treatment
successfully simulates the observed time-dependent inversion
transitions are slower than the D
their spin-forbidden nature.
All of the subsequent absorptive signals for D
and 2 have longer lifetimes than those of the emissive signals,
1 0
f D transitions because of
1 0
, Q, and D in
1
which implies that their observed lifetimes may in fact be limited
by spin relaxation. However, the data in Table 2 show that the
lifetime of 3*PDI (τD2 ) 0.54 µs) for 1 measured in toluene
using nanosecond transient optical absorption agrees reasonably
well with the time constants for the decay of the absorptive
TREPR signal from D (τ ) 620 ns) and Q (τ ) 410 ns),
1
of the initial spin polarization of D and Q within a Zn
tetraphenylporphyrin coordinated to 3-pyridyl-t-Bu-nitroxide.
The authors emphasized that the observed polarization inversion
1
2
2
given the error bars on the TREPR kinetics. Since the optical
kinetic measurements of 3*PDI monitor the decay of the total
1
of both the D and Q signals at later times is essentially a kinetic
effect resulting from the competition between spin selective and
nonselective transition rates.
D and Q population and are not spin selective or sensitive to
spin relaxation, the agreement between the kinetics determined
1
Even though it is difficult to determine a complete set of rate
constants experimentally at room temperature in fluid solution,
a semiquantitative picture of the spin dynamics of these states
within 1-3 can be obtained by examining the kinetic parameters
by optical and TREPR techniques strongly suggests that spin
relaxation contributes little to the overall measured TREPR
kinetics.
A similar picture emerges upon examination of the TREPR
kinetics for 3 except for the fact that since 3J < 0, the energy
that are observed. The appearance of the D
signals in 1-3 all occur within the 30 ns IRF. The spin-allowed
nature of the D f D transitions relative to the D f Q
transitions results in overpopulation of D . As discussed above,
the subsequent transitions D f Q leave |D (+1/2)〉 overpopu-
lated in 1 and 2, which results in an emissive D signal. For
molecule 1, the decay of the D signal from emission to
absorption occurs very rapidly (τ ) 46 ns), so that this decay
monitors the sum of the rates of the allowed D f Q transitions
Figure 6a) and those of D f D . The decay of the emissive
) 140 ns) largely reflects the sum of the rates of
transitions that repopulate D (Figure 6b) and the
transitions
transitions because the Q-D
1
and Q TREPR
ordering of D and Q is reversed, and the temporal order of the
1
2
1
2
absorptive and emissive transitions is inverted. Once again, the
1
spin-allowed nature of the D
f Q transitions results in overpopulation of D
unlike the case of 1 and 2, the subsequent D f Q transitions
leave |D (-1/2)〉, |Q(-3/2)〉, and |Q(-1/2)〉 overpopulated in
3, which produces initially absorptive D and Q signals. The
absorptive signals from both D (τ ) 32 ns) and Q (τ ) 150
2
f D
1
transitions relative to the
1
1
D
2
1
. However,
1
1
1
1
1
1
1
1
1
1
(
1
0
ns) decay with the same time constants as the corresponding
emissive signals of 1 within experimental uncertainty. In
Q signal (τ
the Q f D
1
addition, the appearance time constant for D
also the same as that of 1 within experimental uncertainty, which
0
(τ ) 110 ns) is
1
1
1
Q f D
0
transitions to ground state. The Q f D
1
should be faster than the Q f D
0
0
is once again consistent with the overall rate of the appearance
of the emissive D signal being limited by repopulation of D
0 1
energy gap is much larger than the Q-D
1
gap (cf. eq 10). The
signal because the
Q signal is much stronger than the D
1
1 0
followed by rapid decay of the spin-allowed D f D transitions.
The data in Table 2 show that the lifetime of 3*PDI (τD2 ) 0.80
µs) measured in toluene using nanosecond transient optical
absorption agrees very well with the time constants for the decay
transition probabilities between the spin sublevels of Q are about
2
2
4
times larger than those of D
from return of the system to the ground-state PDI- BPNO is
polarized in the same sense as is D and Q, yet appears with τ
120 ns, which is slower than the IRF. The 120 ns appearance
1
.
0
The D signal that results
2
•
of the absorptive TREPR signal from D
(τ ) 740 ns). Thus, it is likely that the submicrosecond decay
rates for these absorptive signals do not have significant
contributions from spin relaxation. The much longer D decay
time constant of τ > 2 µs that is apparent for 1 is no longer
1
2
(τ ) 800 ns) and Q
1
1
)
2
time constant for the emissive D
0
signal agrees within experi-
mental uncertainty with the 140 ns decay of the emissive signal
of Q, which is consistent with the overall rate of the appearance
0
3
observed for 3 as a result of the lower signal intensity in 3.
For molecule 2, the time constants for the decay of the
of the emissive D
followed by rapid decay of the partially spin-allowed D
transitions. The corresponding Q f D
0
signal being limited by repopulation of D
1
1
f D
0
0
intersystem crossing
emissive D
1
(τ
1
) 200 ns) and Q (τ ) 210 ns) signals to their
1
J. AM. CHEM. SOC. 9 VOL. 131, NO. 10, 2009 3711

A R T I C L E S
Giacobbe et al.
respective absorptive signals, as well as the appearance of the
constants of τ = 2 ps. Rapid photogeneration of the three-spin
emissive D
0
(τ
1
) 200 ns) signal, are all the same within
1
system results in the formation of excited doublet (D ) and
experimental uncertainty. The additional phenyl spacer in 2
should decrease |3J| due to the increased distance between the
quartet (Q) states involving the strongly coupled unpaired
electrons confined to PDI (i.e., 3*PDI), which are each much
2
•
2
•
two unpaired spins on PDI and BPNO and decrease HZFS as a
result of a small increase in the delocalization of the PDI
unpaired electrons onto the phenyl bridge. According to eq 10,
decreasing |3J| reduces the energy gaps between the spin
more weakly coupled to the unpaired electron on BPNO .
2
•
Attachment of BPNO at its 4-position to the imide nitrogen
atom of 3*PDI (1) or by means of an intervening p-phenylene
group (2) results in ferromagnetic coupling (3J > 0) between
2
•
sublevels, which should increase the rates of the D
1
T Q
the spins, while attachment of BPNO at its 3-position to the
imide nitrogen atom of 3*PDI (3) results in antiferromagnetic
coupling (3J < 0) between the spins. Following photoexcitation
of PDI and ultrafast EISC, the TREPR experiments at high
magnetic fields provide sufficient spectral resolution to allow
transitions, while decreasing HZFS should decrease these rates.
Our data suggest that changes in HZFS have slightly more
influence on the measured rates because the decay times of the
1
emissive signals for D and Q are only about 4 and 1.5 times
longer, respectively, than those of 1. The overall slowing of
the spin dynamics within 2 is reflected in the time constant for
the observation of the dynamics of the spin-polarized D
1
and Q
excited states as well as the spin-polarized ground-state of
2
•
the decay of the initially emissive D
0
signal to the absorptive
0
PDI- BPNO (D ). The rapid, photon-controlled formation of
signal, which is also slow and does not invert during the
observed 3 µs time window shown in Figure 4. The decay time
specific polarized spin states that can be achieved by tailoring
the structure of the PDI-radical system gives us an important
building block for assembling a variety of complex multispin
molecular systems in which spin information can be transferred
to specific sites within the systems. This type of behavior has
potential uses in the development of new materials for organic
spintronics.
constants of the absorptive D
1
2
and Q TREPR signals are τ > 3
3*
µs, which is consistent with the PDI lifetime of 2 (τD2 ) 8.4
µs) measured in toluene using nanosecond transient optical
absorption. At these long decay times, spin relaxation most likely
0
contributes to the overall decay of D measured by TREPR.
Finally, one aspect of the RQM mechanism originally
proposed for the TEMPO-C60 system that does not apply to
1 0
-3 relates to the relaxation of D back to the ground-state D .
Acknowledgment. This work was supported by the National
Science Foundation, under Grant No. CHE-0718928, and by the
MRSEC program of the NSF through the Northwestern MRSEC
1
In the TEMPO-C60 system, electron transfer from TEMPO to
60 to produce a CT state followed by charge recombination
leading to the ground state was proposed, D f CT f D . In
(
DMR-0520513). The authors would also like to acknowledge Dr.
C
Z. Dance and Dr. R. Carmieli for their assistance in the TREPR
data collection. M.T.C. thanks the Link Foundation for a Fellowship.
1
0
91
3*
our systems, however, the energy of PDI is only 1.2 eV, so
that even in polar solvents, ∆GCS > 0, and electron transfer
cannot occur.
Supporting Information Available: Experimental details,
including synthesis and additional transient absorption and
TREPR data. This material is available free of charge via the
Internet at http://pubs.acs.org.
Conclusions
Our results show that the BPNO radical quenches 1*PDI
2
•
exclusively by the EISC and EIC mechanisms with time
JA808924F
3
712 J. AM. CHEM. SOC. 9 VOL. 131, NO. 10, 2009