Polytriphenylene dendrimers: A unique design for blue-light-emitting materials

Article abstract of DOI:10.1002/anie.200802854

(Figure Presented) Into the blue: A unique family of polytriphenylene dendrimers (see picture) that emit blue light with high photoluminescence quantum yield were synthesized. The triphenylene units are twisted out-of-plane, and the highly stiff dendritic

Full text of DOI:10.1002/anie.200802854

Communications
DOI: 10.1002/anie.200802854
Dendrimers
Polytriphenylene Dendrimers: A Unique Design for Blue-Light-
Emitting Materials**
Tianshi Qin, Gang Zhou, Horst Scheiber, Roland E. Bauer, Martin Baumgarten,
Christopher E. Anson, Emil J. W. List, and Klaus Mꢀllen*
In the past decade, light-emitting dendritic materials have
attracted much interest, owing to their application as emissive
layers in organic light-emitting diodes (OLEDs).[1] This class
of materials allows the advantages of small molecules, such as
anthracene, phenanthrene, or pyrene[2] to be combined with
those of conjugated polymers, such as polyfluorenes[3] or
poly(para-phenylene vinylene)s.[4] To date, the most promis-
ing design for dendrimer light-emitting-diode (DLED) mate-
rials has comprised a fluorescent[5] or a phosphorescent[6] core
and a conjugated, but non-emissive, dendron shell to keep the
emissive units separated, thus avoiding excimer formation
and quenching effects. The utilization of conjugated dendrons
facilitates sufficient charge transport, which is a necessary
premise for efficient devices. One of the biggest advantages of
using dendritic molecules is the variety of possible combina-
tions of cores, dendrons, and surface groups.[7]
photovoltaic cells.[10] However, to date, although a number of
triphenylene-based small molecules and polymers have been
designed and investigated, the high tendency towards self-
association of triphenylene often leads to a dramatic decrease
in fluorescence intensity, owing to their strong p–p stacking
interactions.[11] To our knowledge, dendrimers based exclu-
sively on interlinked triphenylene units have not been
realized, probably as a result of the challenges associated
with functionalized triphenylene derivatives serving as
branching reagents in dendrimer and hyperbranched polymer
synthesis. In this study, triphenylene is incorporated as a
building block for the dendritic systems G1 (Scheme 1), G2,
The attractive properties of multichromophoric dendrim-
ers[8] suggested a new approach towards blue-light-emitting
materials with the following characteristics: a) blue-light
emission is brought about by the presence of electronically
decoupled polycyclic aromatic hydrocarbon units; b) the units
are incorporated into a rigid dendritic polyphenylene scaf-
fold; they thus adopt sterically defined positions and avoid
intradendrimer chromophore–chromophore interactions;
c) amorphous films are obtained as a result of the lack of
intermolecular interactions, and d) the amount of extraneous
substituents and coupling units is kept at a minimum.
Scheme 1. Synthesis of the polytriphenylene dendrimer G1 (CO is also
released during the reaction). The blue sections indicate the tripheny-
lene units which are the blue-emitting chromphores.
Triphenylene derivatives have been extensively studied
with regard to their discotic liquid crystallinity[9] and opto-
electronic applications in OLEDs, field-effect transistors, and
and G3 (Scheme 2). Though bearing a higher risk of intra-
molecular (triphenylene–triphenylene) chromophore inter-
actions,[12] the use of triphenylene allows for a higher density
of emitting units in the active layer. Therefore, the amount of
electro-inactive material contained in dendrimer films can be
kept at a minimum. Chemical, photophysical, and opto-
electronic characterizations show that the new triphenylene-
based dendritic materials are promising candidates for blue-
light-emitting devices.
The first-generation dendrimer G1 (Scheme 1), with four
triphenylene units, was synthesized by heating the commer-
cially available “phencyclone” 1 and an o-xylene solution of
tetra(4-ethynylphenyl)methane at reflux in a microwave
reactor for 2 h. After the removal of the excess of 1 by
repeated precipitation in methanol, the first-generation
dendrimer G1 was recrystallized from a mixture of tetra-
chloroethane and hexane.
[*] T. Qin, Dr. G. Zhou, Dr. R. E. Bauer, Dr. M. Baumgarten,
Prof. Dr. K. Mꢀllen
Department of Synthetic Chemistry
Max-Planck-Institut fꢀr Polymerforschung
Ackermannweg 10, 55128 Mainz (Germany)
Fax: (+49)6131-379-350
E-mail: muellen@mpip-mainz.mpg.de
H. Scheiber, Prof. Dr. E. J. W. List
Institute of Solid State Physics, Graz University of Technology
Petersgasse 16, 8010 Graz (Austria)
Dr. C. E. Anson
Institut fꢀr Anorganische Chemie, Universitꢁt Karlsruhe
Engesserstrasse 15, 76131 Karlsruhe (Germany)
[**] This work was financially supported by the Deutsche Forschungs-
gemeinschaft (DFG) within the frame of the Sonderforschungs-
bereich (SFB)625. G. Z. gratefully acknowledges the Alexander von
Humboldt Stiftung for a research fellowship.
The synthesis of the higher-generation polytriphenylene
dendrimers G2 and G3 was carried out following a divergent
synthetic protocol.[13] The key step is the utilization of a novel
AB2-type compound, that is, diphenylcyclopentaphenanthre-
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200802854.
8292
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 8292 –8296

Angewandte
Chemie
substituted using ethynyltriiso-
propylsilane under Sonoga-
shira–Hagihara conditions, to
afford
9,10-dimethoxy-3,6-
bis((triisopropylsilyl)ethynyl)-
phenanthrene (4). Upon oxida-
tion with cerium ammonium
nitrate (CAN), the methoxy
groups in 4 were cleaved and
3,6-bis((triisopropylsilyl)ethy-
nyl)phenanthrene-9,10-dione
(5) was obtained. The final step,
by Knoevenagel condensation
of 5 with 1,3-diphenylacetone,
gave the central building block
3,6-bis((triisopropylsilyl)ethy-
nyl)phencyclone 6 (Scheme 3).
Dendrimers G2, with 12 tri-
phenylene units, and G3, with
28 triphenylene units, were syn-
thesized from building block 6
by Diels–Alder [4+2] addition
(Scheme 2). The triisopropyl-
silyl (tips) groups in dendrimers
were deprotected with tetrabu-
tylammonium fluoride (tbaf)
and the resulting ethynyl
groups were further treated
with either building block 1 to
obtain G2, or with 6, continuing
the divergent synthetic method,
to furnish G3. For comparison,
a model compound 1,2,4-tri-
phenyltriphenylene
(ttp;
Scheme 2) containing the tri-
phenylene repeating unit, and
another polyphenylene deriva-
tive, dendrimer G2’ (Scheme 2)
with a pentaphenyl shell, were
also prepared by Diels–Alder
reactions. The structures of
dendrimers G1, G2, and G3
were elucidated by NMR spec-
Scheme 2. Synthesis of polytriphenylene dendrimer G2, G3, G2’, and the model compound 1,2,4-
triphenyltriphenylene (ttp).
none derivative 6 (Scheme 3), which combines a diene
function and two protected ethynyl functions in the same
molecule. Compared to the tetraphenylcyclopentadienone
derivatives used in the synthesis of known polyphenylene
dendrimers,[13] compound 6 contains an additional bond
between two neighboring phenyl rings, a Diels–Alder [4+2]
reaction leads to triphenylene units. The synthesis of the key
intermediate compound 6 began with the reaction of com-
mercially available phenanthrene-9,10-dione with bromine in
nitrobenzene, according to standard methods, to give 3,6-
dibromophenanthrene-9,10-dione (2, Scheme 3), and the
subsequent protection of the dione group in 2 by reductive
methylation, to form 3,6-dibromo-9,10-dimethoxyphenan-
threne (3). The bromo groups in compound 3 were then
Scheme 3. Synthesis of diphenylcyclopentaphenanthrenone 6. CAN=
cerium ammonium nitrate, TEA=triethylamine.
Angew. Chem. Int. Ed. 2008, 47, 8292 –8296
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
8293

Communications
Table 1: Optical data for dendrimers and model compound.
troscopy and MALDI-TOF mass spectrometry. The MALDI-
Abs
max
PL
max
TOF mass spectra each revealed a single intense signal
corresponding to the calculated mass of dendrimers G1, G2,
and G3 (Figure 1).
Sample
l
[nm]
l
[nm]
Quantum yield [%][a]
Solution Thin film Solution Thin film
ttp
G1
G2
G3
G2’
289
283
304
299
312
298
293
312
313
317
401
400
414
424
408
402
406
431
442
418
3.8
6.6
27.0
35.2
14.8
[a] relative to quinine sulfate dehydrate.[16]
Figure 1. MALDI-Mass spectra for G1, G2, and G3.
To investigate the spatial arrangements of polytripheny-
lene dendrimers and their optimum tetrahedral geometry,
single-crystal X-ray analysis of G1 was performed to provide
the most direct description of the molecular structural
features.[14] The single crystal of dendrimer G1 (see Support-
ing Information, Figure S1) was obtained as colorless, tiny,
needle-shaped crystals, by slow evaporation of a solution in a
1:1 mixture of tetrachloroethane and hexane at room temper-
ature (see Supporting Information, Table S1, monoclinic, cell
parameters: a = 17.46, b = 10.81, c = 35.86 ꢀ, and b =
91.08).[15] In the single crystal, each triphenylene unit is
twisted out of the plane, as a result of the attached phenyl
rings. Furthermore, the phenyl substitution reduces the p–p
stacking tendency. In one molecule, the four triphenylene
units are separated by a tetraphenylmethane core, breaking
the conjugation. The molecules are packed in such a way that
the triphenylene branches of neighboring dendrimers are
aligned orthogonal to each other, and the shortest distance
between adjacent triphenylene units is 3.73 ꢀ (see Supporting
Information, Figure S2). Therefore, the triphenylene units are
not in a p-stacked arrangement, which would quench
fluorescence.
The dendrimers G1–G3 are readily soluble in common
solvents, such as chloroform, toluene, and tetrahydrofuran
(THF). Their UV/Vis absorption and photoluminescence
(PL) properties were investigated in toluene solutions
(10ꢀ3 gLꢀ1) and spin-coated films (Table 1). Compounds ttp
and dendrimer G1 demonstrate identical absorption and
photoluminescence maxima as, since the four ttp units in G1
are connected through a sp3 carbon, they have the same
chromophores (Figure 2a). With the elongation of the longest
conjugated oligotriphenylene segments, the absorption and
photoluminescence maxima of dendrimers G2 and G3 are
slightly bathochromically shifted as a consequence of the
increasing number of neighboring ttp units on each branch of
the molecule, in which some p-orbital overlap occurs between
Figure 2. UV/Vis absorption and photoluminescence (PL) spectra
(lex =300 nm) of dendrimers and the model compound in toluene
solution (a) and in the solid state (b).
the ttp units. The intermolecular interactions and aggrega-
tions of dendrimers G1, G2, and G3 were investigated by
measuring the absorption and photoluminescence spectra of
their toluene solutions at different concentrations. Over three
orders of magnitude (10ꢀ4, 10ꢀ3, and 10ꢀ2 gLꢀ1) no bath-
ochromic-shift and excimer emissions could be observed (see
Supporting Information, Figure S3). This concentration inde-
pendence suggests that the dendritic branches in each
dendrimer effectively suppress the intermolecular interac-
tions and prevent aggregation.[16] Absence of aggregation is
also evidenced by the rather high photoluminescence quan-
tum yield (QY) of these dendrimers. With quinine sulfate as a
standard,[17] the quantum yields of dendrimers G1, G2, and
G3 in dilute toluene solutions (10ꢀ6 m) were measured as 6.6,
27.0, and 35.2%, respectively. These values reveal an
increasing quantum yield scaling with the size of the
molecules. This increase could be ascribed to the reduced
branch rotation in higher generations which decreases the
vibrational relaxation and the intersystem crossing in the
excited state. Another plausible explanation for the higher
quantum yield in higher generations could invoke the
8294 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 8292 –8296

Angewandte
Chemie
interactions between the chromophores and environment:
With increasing dendritic generation, the central chromo-
phores in the larger molecules are more effectively shielded,
which suppresses interactions among the chromophores
themselves or between the chromophores and solvent mol-
ecules, which would otherwise result in fluorescence quench-
ing.
Figure 2b shows the absorption and photoluminescence
spectra of dendrimers G1, G2, and G3 in solid states. In the
photoluminescence spectra of ttp and G1, bathochromic shifts
of 1 nm and 6 nm, respectively, were observed compared to
those of the solutions. However, dendrimers G2 and G3
display a more pronounced bathochromic shift of 17–18 nm,
respectively, compared to their corresponding solution spec-
tra. We attribute this bathochromic shift to the solid-state
packing which leads to increased coupling of individual ttp
units with increasing generations. It is notable that the
absorption happens mainly from S0!S3 and S0!S4,[11,18]
because S0!S1 and S0!S2 are symmetry forbidden in isolated
triphenylene molecules and only unresolved shoulder peaks
can be observed around 350 nm for these transitions.[11] The
broadening of the high-wavelength flank of absorption peaks
could be due to reduced symmetry[19] caused by the non-
planarity, attached phenyl rings, and neighboring tripheny-
lene units, thus increasing the oscillator strength of the S0!S1
and S0!S2 transitions.
Figure 3. I–V–L characteristics of an ITO/PEDOT:PSS/G2/tpbi(10 nm)/
CsF/Al device. Inset: normalized electroluminescence (EL) spectrum at
8 V bias.
dendritic system, which indicates that the dendritic branches
efficiently suppress the intermolecular interaction. The device
displayed an onset of electroluminescence at approximately
6 V and maximum efficiency of 0.4 cdAꢀ1. According to the
energy level diagram (see Supporting Information, Fig-
ure S6), the rather poor efficiency is most probably attributed
to the large electron-injection barrier from CsF (ꢀ4.1 eV) to
tpbi (ꢀ2.8 eV). This large barrier also explains the relatively
high turn-on voltage of this device. Therefore, the device
efficiency could be improved by choosing appropriate cath-
odes and electron transporting materials. Investigations into
device optimization are underway.
For comparison, the performance of the DLED based on
dendrimer G2’ was measured under identical conditions (see
Supporting Information, Figure S7). The normalized electro-
luminescence spectrum resembles the thin film photolumi-
nescence spectrum very well and is located in a deeper blue
region than G2 with a maximum at 415 nm and corresponding
CIE coordinates of (0.17, 0.10). At 10 V driving voltage, a
maximum luminance of 400 cdmꢀ2 was found with an
efficiency of 0.1 cdAꢀ1. Compared to G2, this lower efficiency
is mainly due to the relatively lower quantum yield found for
photoluminescence (Table 1). Another possibility may be that
a significant part of the electroluminescence spectrum of G2’,
that from the pentaphenyl shell, is in the UV region and
therefore does not contribute to the luminance value.
Overall it becomes clear that the performance of the
presented devices can compete with the best reported
fluorescence-based blue-emitting DLEDs with respect to
device efficiency and brightness,[1,22] which also holds true for
a comparison with fluorescent blue-light-emitting polymeric
devices based on poly(para-phenylene)-type polymers.[23]
Utilizing transport moieties in the outer shell of the dendri-
mer and tuning of the emission color more from the UV to the
blue region, both strategies that are successfully implemented
in PLED materials,[23] will further allow for improvement of
device performance in fluorescence-based blue-emitting
DLEDs.
The thermal properties of dendrimers G1, G2, and G3
were studied by thermogravimetric analysis (TGA, see
Supporting Information, Figure S4). In a nitrogen atmos-
phere, they exhibit degradation above 4508C, which is similar
to other reported polyphenylene dendrimers.[20] The photo-
luminescence spectra of dendrimers G2 and G2’ before and
after annealing at 2008C show no significant changes, even
after annealing in air (Figure S5 in the Supporting Informa-
tion). The above measurements demonstrate the high thermal
stability of these dendrimers, making them promising materi-
als for DLEDs.
To test the electroluminescent properties of dendrimers
G2 and G2’, DLEDs were fabricated in a standard sandwich
geometry using the following configuration: indium tin oxide
(ITO)/ poly(styrene sulfonate)-doped poly(3,4-ethylenedioxy-
thiophene) (PEDOT:PSS)/G2 (or G2’)/1,3,5-tris(1-phenyl-
1H-2-benzimidazolyl)benzene (tpbi)/CsF/Al. To avoid recom-
bination at the chemically unstable cathode interface,[21] tpbi
was applied as an additional electron transport and hole-
blocking layer.
Figure 3 shows the current–voltage (I–V) and luminance–
voltage (L–V) characteristics of a DLED with dendrimer G2
as the emitting layer. The device emitted a sky-blue electro-
luminescence (EL) with a maximum brightness of 300 cdmꢀ2
at a bias voltage of 8 V and corresponding Commission
Internationale de LꢁEclairage (CIE) coordinates of (0.19,
0.18). The maximum of the electroluminescence spectrum
peaked at 430 nm. Both the maximum and the shape of the
electroluminescence spectrum are similar to the correspond-
ing photoluminescence spectrum of a G2 film. Excimer
emission from chromophore aggregation and keto-defect
emission from oxidative degradation, which were always
found for blue LEDs, however, could not be observed in our
In conclusion, we have synthesized a series of blue light-
emitting dendrimers by a unique procedure. The synthesis of
Angew. Chem. Int. Ed. 2008, 47, 8292 –8296
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
8295

Communications
[5] M. Halim, J. N. G. Pillow, P. L. Burn, I. D. W. Samuel, Adv.
Mater. 1999, 11, 371.
[6] S.-C. Lo, E. B. Namdas, P. L. Burn, I. D. W. Samuel, Macro-
molecules 2003, 36, 9721.
[7] J. N. G. Pillow, M. Halim, J. M. Lupton, P. L. Burn, I. D. W.
Samuel, Macromolecules 1999, 32, 5985.
[8] J. Luo, Y. Zhou, Z. Q. Niu, Q. F. Zhou, Y. Ma, J. Pei, J. Am.
Chem. Soc. 2007, 129, 11314.
[9] a) S. Kumar, Liq. Cryst. 2005, 32, 1089; b) S. Kumar, Liq. Cryst.
2004, 31, 1037.
[10] K. H. Li, Z. T. Xu, H. H. Xu, J. M. Ryan, Chem. Mater. 2005, 17,
4426.
[11] V. Duzhko, H. F. Shi, K. D. Singer, A. N. Semyonov, R. J. Twieg,
Langmuir 2006, 22, 7947.
[12] T. W. Kwon, M. M. Alam, S. A. Jenekhe, Chem. Mater. 2004, 16,
4657.
polytriphenylene dendrimers starting from AB2-type building
blocks is not catalyst driven and gives essentially quantitative
yield. Most importantly, the twisted triphenylene units not
only function as chromophores, but also effectively prevent
inter- or intramolecular fluorescence quenching, as demon-
strated by the increase in photoluminescence quantum yield
with increasing generation. These dendrimers exhibit stable
and pure-blue emission in both photoluminescence and
electroluminescence spectra and could provide an avenue
for dendritic emitters with the optimized electroluminescence
efficiency/color purity compromise needed for pure blue
light, which is still a great challenge in the field of blue
emitters.
[13] U. M. Wiesler, A. J. Berresheim, F. Morgenroth, G. Lieser, K.
Mꢂllen, Macromolecules 2001, 34, 187.
Received: June 16, 2008
Published online: September 18, 2008
[14] R. E. Bauer, V. Enkelmann, U. M. Wiesler, A. J. Berresheim, K.
Mꢂllen, Chem. Eur. J. 2002, 8, 3858.
[15] CCDC 691521 contains the supplementary crystallographic data
for this paper. These data can be obtained free of charge from
The Cambridge Crystallographic Data Centre via www.ccdc.
cam.ac.uk/data_request/cif.
Keywords: chromophores · dendrimers · fluorescence ·
luminescence · oleds
.
[1] a) C. D. Mꢂller, A. Falcou, N. Reckefuss, M. Rojahn, V.
Wiederhirn, P. Rudati, H. Frohne, O. Nuyken, H. Becker, K.
Meerholz, Nature 2003, 421, 829; b) C. Ma, E. Mena-Osteritz, T.
Debaerdemaeker, M. M. Wienk, R. A. J. Janssen, P. Bꢃuerle.
Angew. Chem. 2007, 119, 1709; Angew. Chem. Int. Ed. 2007, 46,
1679; Angew. Chem. Int. Ed. 2007, 46, 1679; c) R. H. Friend,
R. W. Gymer, A. B. Holmes, J. H. Burroughes, R. N. Marks, C.
Taliani, D. D. C. Bradley, D. A. Dos Santos, J. L. Brꢄdas, M.
Lꢅgdlund, W. R. Salaneck, Nature 1999, 397, 121.
[16] D. Markovitsi, A. Germain, P. Milliꢄ, P. Lꢄcuyert, L. K. Gallos, P.
Argyrakis, H. Bengs, H. Ringsdorf, J. Phys. Chem. 1995, 99, 1005.
[17] A. N. Fletcher, Photochem. Photobiol. 1969, 9, 439.
[18] H. Chojnacki, Z. Laskowski, A. Lewanowicz, Z Ruziewicz, R.
Wandas, Chem. Phys. Lett. 1986, 124, 478.
[19] S. Marguet, D. Markovitsi, P. Milliꢄ, H. Sigal, S. Kumar, J. Phys.
Chem. B 1998, 102, 4697.
[20] V. U. M. Wiesler, T. Weil, K. Mꢂllen, Top. Curr. Chem. 2001, 212,
1.
[2] a) Y. H. Kim, H. C. Jeong, S. H. Kim, K. Yang, S. K. Kwon, Adv.
Funct. Mater. 2005, 15, 1799; b) M. J. D. Bosdet, C. A. Jaska,
W. E. Piers, T. S. Sorensen, M. Parvez, Org. Lett. 2007, 9, 1395;
c) W. L. Jia, T. McCormick, Q. D. Liu, H. Fukutani, M. Motala,
R. Y. Wang, Y. Tao, S. N. Wang, J. Mater. Chem. 2004, 14, 3344.
[3] J. M. Lupton, Nature 2008, 453, 459.
[4] J. H. Burroughes, D. D. C. Bradley, A. R. Brown, R. N. Marks,
K. Mackay, R. H. Friend, P. L. Burns, A. B. Holmes, Nature 1990,
347, 539.
[21] S. Gamerith, H.-G. Nothofer, U. Scherf, E. J. W. List, Jpn. J.
Appl. Phys. 2004, 43, 891.
[22] J. P. J. Markham, E. B. Namdas, T. D. Anthopoulos, I. D. W.
Samuel, G. J. Richards, P. L. Burn, Appl. Phys. Lett. 2004, 85,
1463.
[23] a) J. Jacob, A. C. Grimsdale, K. Mꢂllen, S. Sax, M. Gaal, E. J. W.
List, Macromolecules 2005, 38, 9933; b) A. Pogantsch, F. P.
Wenzl, E. J. W. List, G. Leising, A. C. Grimsdale, K. Mꢂllen,
Adv. Mater. 2002, 14, 1061.
8296 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 8292 –8296