Tuning the singlet-triplet gap in metal-free phosphorescent π-conjugated polymers

Article abstract of DOI:10.1002/anie.201003291

Glow in the dark: A tunable singlet-triplet exchange gap arises in optically active conjugated polymers that exhibit both strong fluorescence from the singlet state and pronounced phosphorescence from triplet excitations (see picture). The polymers were incorporated into organic light-emitting diodes (OLEDs).

Full text of DOI:10.1002/anie.201003291

Communications
DOI: 10.1002/anie.201003291
p-Conjugated Polymers
Tuning the Singlet–Triplet Gap in Metal-Free Phosphorescent
p-Conjugated Polymers**
Debangshu Chaudhuri, Henning Wettach, Kipp J. van Schooten, Su Liu, Eva Sigmund,
Sigurd Hꢀger,* and John M. Lupton*
Much of the fascination in conjugated polymers stems from
the realization of potential device applications,[1] yet the true
power of these materials lies in the possibility to synthesize a
compound with a particular property from first principles. A
defining feature of these carbon-based semiconductors is that
spin–orbit coupling is weak, thus making spin states well-
defined.[2–3] Electrical injection of charges into an organic
light-emitting diode (OLED) leads to triplet and singlet
excitations with symmetric or antisymmetric spin wavefunc-
tions. As triplet excitations are typically not dipole-coupled to
the molecular ground state, the majority of carrier pairs that
recombine in an OLED decay nonradiatively. Incorporation
of heavy-atom centers that are mostly in the form of an
organometallic luminophore and promote spin–orbit cou-
pling, is employed to promote radiative triplet recombina-
tion.[4]
The direct spectroscopic identification of triplet excita-
tions in conjugated polymers became possible by detecting
the weak phosphorescence.[5] In this approach, the relative
energies of the singlet and triplet excitations are determined
by comparing prompt fluorescence to delayed phosphores-
cence, thus directly revealing the exchange energy. Although
small-molecule organometallic complexes display a variety of
exchange interaction strengths,[4e,f] it has been proposed that
the singlet and triplet excited state are universally split by
some 0.7 eV in conjugated polymers.[3c] Here, we present a
new series of triphenylene-based metal-free conjugated
copolymers with tunable exchange splitting, which is revealed
by their distinct fluorescence and phosphorescence signa-
tures. In addition, we show how both singlet and triplet
excited states can couple to emissive defects, hence revealing
the fundamental mechanism for color purity degradation that
is a common problem for poly(para-phenylene)s such as
polyfluorenes.[6]
Most conjugated hydrocarbons show either weak phos-
phorescence or no phosphorescence at all.[3] Triphenylenes,
and some other polycyclic aromatic hydrocarbons, are an
exception: at low temperatures, they exhibit a pronounced
afterglow, which is often attributed to the phosphorescence
from the triplet state.[7] Although triphenylene-based com-
pounds have recently been investigated as blue singlet
emitters in OLEDs,[8] the relevance of the triplet states to
OLEDs has not yet been considered.
In the polymers presented herein, the triphenylene 1 is an
integral part of the polymer backbone (Figure 1). Incorpo-
ration of 1 into the backbone leads to a fully conjugated
p system in which the triplet state is localized on the
triphenylene unit, yet the singlet state is delocalized over
multiple repeat units (Figure 1a). We synthesized the poly-
mer 2 from 1,[9] along with four different copolymers 3–6 by
using transition-metal-catalyzed polycondensation. The
prompt and delayed luminescence of the compounds dis-
persed at a 1% weight ratio in polystyrene at 25 K is shown in
Figure 1b–g. The prompt luminescence (detected in a 2 ns
time window that coincides with the laser pulse) is caused by
fluorescence from the singlet state, whereas the delayed
emission (detected several hundred microseconds after exci-
tation) can be assigned to phosphorescence from the triplet
state.[7] A similar triplet spectrum is observed for all materials
(the average peak position is marked by a dashed line),
whereas the singlet peak strongly depends on the intricacies
of the polymer backbone.[9] The energetic separation between
fluorescence and phosphorescence provides an accurate
measure of the magnitude of the exchange interaction
(2J).[3] From 1 to 6, the singlet emission shifts to the red,
while the triplet emission remains unchanged:[9] copolymer 6
has almost degenerate singlet and triplet levels. This degen-
eracy is unprecedented in conjugated polymers and is
particularly surprising given the fact that the triplet level of
regular poly(thienylene vinylene) lies at 1170 nm.[3i] Appa-
rently, singlet and triplet excitations can form on different
parts of the conjugated system,[10] thus allowing the splitting to
be tuned. The series in Figure 1 suggests that it may be
possible to design materials in which the regular level
ordering is reversed, so that the singlet state lies energetically
beneath the triplet state.
[*] Dr. D. Chaudhuri, K. J. van Schooten, S. Liu, Prof. Dr. J. M. Lupton
Department of Physics and Astronomy
University of Utah
Salt Lake City, UT 84112 (USA)
Fax: (+1)801-581-4801
E-mail: lupton@physics.utah.edu
Dr. H. Wettach, E. Sigmund, Prof. Dr. S. Hꢀger
Kekulꢁ-Institut fꢂr Organische Chemie und Biochemie
der Universitꢃt Bonn
Gerhard-Domagk-Str. 1, 53121 Bonn (Germany)
Fax: (+49)228-73-5662
E-mail: hoeger@uni-bonn.de
[**] We thank the Volkswagen Foundation for collaborative research
funding. J.M.L. is a David and Lucile Packard Foundation Fellow and
is indebted to the NSF (grant CHE-ASC 0748473) and the DoE
(grant DESC0000909) for financial support.
The singlet and triplet states of the polymer are not the
only spectral signatures seen in 1–6. Poly(para-phenylene)s
are prone to the formation of oxidative defects that result in
broad emission spectra, as has been studied in particular
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201003291.
7714
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 7714 –7717

Angewandte
Chemie
Figure 2. Singlet and triplet excited states that feed the defect states in
the triphenylene homopolymer 2 at 25 K. a) Prompt singlet emission
gives way to b) a broad defect (X) emission detected in a time window
of 2–4 ns after excitation. c) Similarly, long-lived triplet states (detected
30–80 ms after excitation) transfer to the d) defect states (X’). Subse-
quent emission is seen as an afterglow, which can persist for seconds
after excitation.
band in polyfluorene.[6,11] This species is populated by energy
transfer: the singlet feeds into a trap (X). Fluorenone-type
defect excitations have a charge-transfer character and are
less strongly coupled to the ground state than the singlet,
hence leading to an increased emission lifetime.[6] Once the
species X has decayed, the phosphorescence becomes visible
in a time window of 30–80 ms (c) and lasts for several
milliseconds (Figure 1). The triplet peak is accompanied by a
residual singlet emission around 400 nm that results either
from triplet–triplet annihilation[5] or from the delayed recom-
bination of optically-generated carrier pairs.[12b] Subsequently,
the defect spectrum reappears at very long detection times:
the triplet also feeds into a trap denoted X’, which has
an emission spectrum slightly red-shifted relative to that of
X (d).
Figure 1. Fluorescence (prompt emission) and phosphorescence
(delayed emission) from a triphenylene-based monomer and conju-
gated copolymers, dispersed in a polystyrene matrix at 25 K. a) Deloc-
alization of singlet excitations (blue) with triplets (red) localized at the
triphenylene unit. b)–g) Singlet (solid blue line, integrated 0-2 ns after
excitation) and triplet (dashed red line) spectra of the monomer 1
(0.1–1.1 ms delay after excitation), the homopolymer 2 (9–10 ms
delay), the para-phenylene copolymer 3 (1–2 ms delay), the ethynylene
copolymer 4 (1.5–2.5 ms delay), phenylene vinylene copolymer 5
(0.02–1.02 ms delay), and the thienylene vinylene copolymer 6 (0.05–
5.05 ms delay). The exchange splitting 2J is estimated from the peak
separations. The dashed black line indicates the average triplet peak
position.
To test for electrophosphorescence, we fabricated
OLEDs, in which the majority excitations are triplets. In
contrast to polymers that contain trace concentrations of
metal centers, which locally promote spin–orbit coupling and
thereby enable triplet-diffusion-driven phosphorescence,[12]
we have not yet observed any direct emission from the
triphenylene triplet, even at low temperatures. Figure 3 shows
a comparison of films of the homopolymer 2 under optical
detail in the case of polyfluorene.[6b] We are now able to
elucidate the interplay between triplet excitations and
emissive defects as a distinct phosphorescence signature can
be induced in the polymer emission. A time-dependent series
of normalized emission spectra of 2 is shown in Figure 2. The
prompt singlet emission (a) decays within 2 ns, to give a
broad, featureless spectrum (b) reminiscent of the fluorenone
Angew. Chem. Int. Ed. 2010, 49, 7714 –7717
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7715

Communications
Figure 3. Defect emission and afterglow in triphenylene materials.
Comparison of electrical (black) and optical (red) excitation of the
homopolymer 2 at room temperature. Electroluminescence is domi-
nated by defect emission, which is similar to the afterglow in the
photoluminescence of triphenylene 1 at 25 K, detected at 1 s after
excitation (green).
and electrical excitation. In the OLED, an additional broad
band appears at 540 nm, which is reminiscent of the keto
[6]
defect in polyfluorenes and other poly(para-phenylene)s
and has been observed in other triphenylene-based OLEDs.[8]
This band is not visible under steady-state optical excitation.
The green curve in Figure 3 shows the afterglow photo-
luminescence spectrum of 1 (recorded 1 s after excitation),
which is remarkably similar in shape and position to the
540 nm defect band seen in OLEDs of 2, but is clearly distinct
from the triplet spectrum shown in Figure 1. The observation
of luminescence 1 s after excitation (i.e., X’ emission) requires
an extremely long-lived state. We thus conclude that triplet
states feed directly into defect states, provided that there is an
appropriate spatial overlap of the molecular excited-state
wavefunctions.
The compounds presented herein allow a systematic
investigation of the energetic prerequisites for singlet or
triplet energy transfer to the defect to occur. Generally, the
presence of a single excited state decay channel results in
exponential luminescence kinetics. However, singlet energy
transfer to the species X opens a second decay pathway, thus
leading to nonexponential kinetics in the ensemble of
molecules. Transient luminescence from solutions of polymers
2–6 in CHCl3 are shown in Figure 4a. Whereas the low-J
materials 4–6 show a single exponential decay, materials 2–3
are characterized by nonexponential kinetics. This deviation
from exponential decay correlates with the appearance of a
secondary emissive species (X) within 3 ns after excitation in
polymers 2 and 3 (Figure 4b–f). The species X is much weaker
in 4 and cannot be detected in 5–6. We conclude that the
polymer singlet state lies above the defect absorption in 2–3,
so that energy transfer can occur. The picosecond fluores-
cence dynamics also relate to the afterglow of the materials
recorded 1 s after excitation (Figure 4g–k). Whereas the
afterglow is dominated by the defect in 2–3, the narrow triplet
features are discernible in 4–6: triplet transfer to the defect is
impeded for these materials with smaller J value. Conse-
quently, we found that OLEDs made from 5 showed a much
reduced defect emission with a closer correspondence of
emission under optical and electrical excitation. We therefore
propose that lowering the exchange gap also reduces the
Figure 4. Comparison of singlet and triplet luminescence dynamics of
polymers 2–6. a) Room temperature decay of the singlet emission
intensity in CHCl3. b)–f) Change of the emission spectra of 2–6 within
4 ns of excitation (in CHCl3 at room temperature). Solid lines represent
prompt fluorescence (detected in the first 100 ps) and the dashed
lines illustrate the delayed spectra (within 3–4 ns after excitation). g)–
k) Afterglow spectra at 25 K (in polystyrene films), 1 s after excitation
(solid line), compared to the corresponding phosphorescence spec-
trum (dashed line, from Figure 1).
tendency of triplet excitations to migrate to emissive defects,
which may help in controlling color purity in OLED
materials, and could ultimately enable the construction of
all-organic phosphorescent OLEDs.
The ability to engineer the singlet–triplet gap in con-
jugated polymers is likely to be useful for the control of spin
excitations in organic devices, such as in spin valves and
magnetoresistive magnetic field sensors,[13] because spin
manipulation by magnetic or electric fields in organic semi-
conductors is hindered by the large excitonic exchange
interaction.[12b] Our approach shows that this interaction can
be reduced to some 20 meV. The tuning of the relative
singlet–triplet energies may offer a route to the minimization
of triplet gain quenching in laser structures.[14] Such an
approach could allow a realistic design of all-injection laser
diodes by employing the polymer as a triplet scavenger itself
7716
www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 7714 –7717

Angewandte
Chemie
Galbrecht, D. Q. Gao, U. Scherf, Adv. Funct. Mater. 2006, 16,
in analogy to nonphotochemical quenching in photosynthe-
sis.[15] Finally, polycyclic aromatic hydrocarbons such as
triphenylenes could offer elegant access to the tuning of
hyperfine field effects (which control spin-dependent pro-
cesses[16]), such as by deuteration[13b] or purification of carbon
isotopes. This precise engineering of intramolecular exchange
interactions is thus likely to further broaden the appeal of
conjugated polymers for spintronics applications,[13] and may
even be applicable to problems as far afield as quantum
information processing.[17]
364; h) A. Satrijo, S. E. Kooi, T. M. Swager, Macromolecules
2007, 40, 8833; i) L. R. Faulkner, A. J. Bard, J. Am. Chem. Soc.
1968, 90, 6284.
[7] a) J. Schmidt, W. G. Vandorp, J. H. van der Waals, Chem. Phys.
Lett. 1971, 8, 345; b) J. Langelaar, R. P. H. Rettschnick, A. M. F.
Lamrooy, G. J. Hoytink, Chem. Phys. Lett. 1968, 1, 609; c) N.
Nishi, K. Matsui, M. Kinoshita, J. Higuchi, Mol. Phys. 1979, 38, 1;
d) D. A. Antheunis, J. Schmidt, J. H. van der Waals, Chem. Phys.
Lett. 1970, 6, 255; e) D. Baunsgaard, M. El Balsami, J. Freder-
iksen, N. Harrit, F. Negri, G. Orlandi, R. Wilbrandt, Laser Chem.
1999, 19, 349; f) W. D. K. Clark, A. D. Litt, C. Steel, J. Am.
Chem. Soc. 1969, 91, 5413; g) R. E. Kellogg, R. G. Bennett,
J. Chem. Phys. 1964, 41, 3042; h) A. A. Lamola, G. S. Hammond,
J. Chem. Phys. 1965, 43, 2129.
Received: May 31, 2010
Revised: July 26, 2010
Published online: September 10, 2010
[8] a) M. Saleh, Y. S. Park, M. Baumgarten, J. J. Kim, K. Mꢅllen,
Macromol. Rapid Commun. 2009, 30, 1279; b) A. Rose, C. G.
Lugmair, T. M. Swager, J. Am. Chem. Soc. 2001, 123, 11298;
c) M. Saleh, M. Baumgarten, A. Mavrinskiy, T. Schꢁfer, K.
Mꢅllen, Macromolecules 2010, 43, 137; d) T. S. Qin, G. Zhou, H.
Scheiber, R. E. Bauer, M. Baumgarten, C. E. Anson, E. J. W.
List, K. Mꢅllen, Angew. Chem. 2008, 120, 8416; Angew. Chem.
Int. Ed. 2008, 47, 8292; e) B. N. Boden, K. J. Jardine, A. C. W.
Leung, M. J. MacLachlan, Org. Lett. 2006, 8, 1855; f) H.
Wettach, S. S. Jester, A. Colsmann, U. Lemmer, N. Rehmann,
K. Meerholz, S. Hꢀger, Synth. Met. 2010, 160, 691.
[9] Conjugation of the building blocks is confirmed by considering
an oligomer series of 2, which reveals a bathochromic shift in
fluorescence and absorption. No lower-lying triplet emissions
are seen in any of the low-exchange-gap samples. See the
Supporting Information for further details and spectra.
[10] a) T. Dutta, K. B. Woody, S. R. Parkin, M. D. Watson, J.
Gierschner, J. Am. Chem. Soc. 2009, 131, 17321; b) K. Glusac,
M. E. Kose, H. Jiang, K. S. Schanze, J. Phys. Chem. B 2007, 111,
929.
[11] We generally observe brighter X and Xꢄ emissions in samples
exposed to air, thus suggesting that the defects are a result of
triphenylene oxidation. Elucidation of the actual chemical
structure that is responsible for the defect emission is, however,
complicated by the fact that a small number of defects can be
populated efficiently by energy transfer,[6] and concentrations
are thus likely not very high. We note that some similarity may
exist to the well-documented formation of emissive enol groups
in anthracene electrochemiluminescence.[6i] A complete under-
standing of the present traps will presumably require detailed
computational modeling.
Keywords: fluorescence · organic light-emitting diodes ·
phosphorescence · polymers · triplet excited states
.
[1] A. C. Grimsdale, K. L. Chan, R. E. Martin, P. G. Jokisz, A. B.
Holmes, Chem. Rev. 2009, 109, 897.
[2] a) A. P. Monkman, H. D. Burrows, L. J. Hartwell, L. E. Hors-
burgh, I. Hamblett, S. Navaratnam, Phys. Rev. Lett. 2001, 86,
1358; b) A. P. Monkman, H. D. Burrows, I. Hamblett, S.
Navarathnam, M. Svensson, M. R. Andersson, J. Chem. Phys.
2001, 115, 9046; c) L. S. Swanson, J. Shinar, K. Yoshino, Phys.
Rev. Lett. 1990, 65, 1140; d) A. J. Cadby, C. Yang, S. Holdcroft,
D. D. C. Bradley, P. A. Lane, Adv. Mater. 2002, 14, 57.
[3] a) J. S. Wilson, A. Kꢀhler, R. H. Friend, M. K. Al-Suti, M. R. A.
Al-Mandhary, M. S. Khan, P. R. Raithby, J. Chem. Phys. 2000,
113, 7627; b) D. Wasserberg, P. Marsal, S. C. J. Meskers, R. A. J.
Janssen, D. Beljonne, J. Phys. Chem. B 2005, 109, 4410; c) A.
Kꢀhler, D. Beljonne, Adv. Funct. Mater. 2004, 14, 11; d) A.
Kꢀhler, J. S. Wilson, R. H. Friend, Adv. Mater. 2002, 14, 701;
e) A. Kꢀhler, J. S. Wilson, R. H. Friend, M. K. Al-Suti, M. S.
Khan, A. Gerhard, H. Bꢁssler, J. Chem. Phys. 2002, 116, 9457;
f) D. Beljonne, Z. Shuai, G. Pourtois, J. L. Brꢂdas, J. Phys. Chem.
A 2001, 105, 3899; g) P. C. Jha, E. Jansson, H. ꢃgren, Chem.
Phys. Lett. 2006, 424, 23; h) L. P. Chen, L. Y. Zhu, Z. G. Shuai,
J. Phys. Chem.
A 2006, 110, 13349; i) J. J. Apperloo, C.
Martineau, P. A. van Hal, J. Roncali, R. A. J. Janssen, J. Phys.
Chem. A 2002, 106, 21; j) F. Laquai, C. Im, A. Kadashchuk, H.
Bꢁssler, Chem. Phys. Lett. 2003, 375, 286.
[4] a) M. A. Baldo, D. F. OꢄBrien, Y. You, A. Shoustikov, S. Sibley,
M. E. Thompson, S. R. Forrest, Nature 1998, 395, 151; b) V.
Cleave, G. Yahioglu, P. Le Barny, R. H. Friend, N. Tessler, Adv.
Mater. 1999, 11, 285; c) X. H. Yang, D. C. Mꢅller, D. Neher, K.
Meerholz, Adv. Mater. 2006, 18, 948; d) L. Ying, A. Q. Zhang, W.
Yang, Y. Cao, Prog. Chem. 2009, 21, 1275; e) S. Haneder, E. Da
Como, J. Feldmann, J. M. Lupton, C. Lennartz, P. Erk, E. Fuchs,
O. Molt, I. Mꢅnster, C. Schildknecht, G. Wagenblast, Adv. Mater.
2008, 20, 3325; f) H. Yersin, Top. Curr. Chem. 2004, 241, 1.
[5] a) Y. V. Romanovskii, A. Gerhard, B. Schweitzer, U. Scherf, R. I.
Personov, H. Bꢁssler, Phys. Rev. Lett. 2000, 84, 1027; b) A.
Hayer, H. Bꢁssler, B. Falk, S. Schrader, J. Phys. Chem. A 2002,
106, 11045; c) D. Hertel, S. Setayesh, H. G. Nothofer, U. Scherf,
K. Mꢅllen, H. Bꢁssler, Adv. Mater. 2001, 13, 65.
[6] a) A. P. Kulkarni, X. Kong, S. A. Jenekhe, J. Phys. Chem. B 2004,
108, 8689; b) E. J. W. List, R. Guentner, P. S. de Freitas, U.
Scherf, Adv. Mater. 2002, 14, 374; c) J. M. Lupton, Chem. Phys.
Lett. 2002, 365, 366; d) J. M. Lupton, M. R. Craig, E. W. Meijer,
Appl. Phys. Lett. 2002, 80, 4489; e) J. M. Lupton, J. Klein, Phys.
Rev. B 2002, 65, 193202; f) J. M. Lupton, P. Schouwink, P. E.
Keivanidis, A. C. Grimsdale, K. Mꢅllen, Adv. Funct. Mater. 2003,
13, 154; g) K. Becker, J. M. Lupton, J. Feldmann, B. S. Nehls, F.
[12] a) J. M. Lupton, A. Pogantsch, T. Piok, E. J. W. List, S. Patil, U.
Scherf, Phys. Rev. Lett. 2002, 89, 167401; b) M. Reufer, M. J.
Walter, P. G. Lagoudakis, A. B. Hummel, J. S. Kolb, H. G.
Roskos, U. Scherf, J. M. Lupton, Nat. Mater. 2005, 4, 340;
c) W. Y. Wong, L. Liu, J. X. Shi, Angew. Chem. 2003, 115, 4198;
Angew. Chem. Int. Ed. 2003, 42, 4064.
[13] a) V. A. Dediu, L. E. Hueso, I. Bergenti, C. Taliani, Nat. Mater.
2009, 8, 707; b) T. D. Nguyen, G. Hukic-Markosian, F. Wang, L.
Wojcik, X.-G. Li, E. Ehrenfreund, Z. V. Vardeny, Nat. Mater.
2010, 9, 345.
[14] a) N. C. Giebink, S. R. Forrest, Phys. Rev. B 2009, 79, 073302;
b) J. M. Lupton, Nature 2008, 453, 459.
[15] D. Gust, T. A. Moore, A. L. Moore, A. A. Krasnovsky, P. A.
Liddell, D. Nicodem, J. M. Degraziano, P. Kerrigan, L. R.
Makings, P. J. Pessiki, J. Am. Chem. Soc. 1993, 115, 5684.
[16] a) M. Wohlgenannt, K. Tandon, S. Mazumdar, S. Ramasesha,
Z. V. Vardeny, Nature 2001, 409, 494; b) D. R. McCamey, H. A.
Seipel, S. Y. Paik, M. J. Walter, N. J. Borys, J. M. Lupton, C.
Boehme, Nat. Mater. 2008, 7, 723.
[17] C. Barthel, D. J. Reilly, C. M. Marcus, M. P. Hanson, A. C.
Gossard, Phys. Rev. Lett. 2009, 103, 160503.
Angew. Chem. Int. Ed. 2010, 49, 7714 –7717
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7717