Donor-substituted 1,3,5-triazines as host materials for blue phosphorescent organic light-emitting diodes

Article abstract of DOI:10.1021/cm9033879

A series of new donor-substituted 1,3,5-triazines (TRZ 1-7) has been prepared by nucleophilic substitution of cyanuric chloride with carbazole, 3-methylcarbazole, phenol, and 3,5-dimethylphenol. These s-triazines have been investigated as host material fo

Full text of DOI:10.1021/cm9033879

Chem. Mater. 2010, 22, 2403–2410 2403
DOI:10.1021/cm9033879
Donor-Substituted 1,3,5-Triazines as Host Materials for Blue
Phosphorescent Organic Light-Emitting Diodes
Michael M. Rothmann,† Stephan Haneder,‡ Enrico Da Como,‡ Christian Lennartz,§
Christian Schildknecht,§ and Peter Strohriegl*,†
†Makromolekulare Chemie I and Bayreuther Institut fu€r Makromoleku€lforschung, University of Bayreuth,
95440 Bayreuth, Germany, ‡Photonics and Optoelectronics Group, Department of Physics and CeNS,
Ludwig-Maximilians-Universita€t, 80799 Munich, Germany, and BASF SE, 67056 Ludwigshafen, Germany
§
Received November 9, 2009. Revised Manuscript Received February 18, 2010
A series of new donor-substituted 1,3,5-triazines (TRZ 1-7) has been prepared by nucleophilic
substitution of cyanuric chloride with carbazole, 3-methylcarbazole, phenol, and 3,5-dimethyl-
phenol. These s-triazines have been investigated as host material for blue phosphorescent light-
emitting diodes (OLEDs). All triazine based hosts were characterized regarding their optical and
thermal properties. Different substitution patterns resulted in high glass-transition temperatures (Tg)
of up to 170 °C and triplet energies (ΔE(T1-S0)) of up to 2.96 eV. The application as host material for
the blue phosphor bis(4,6-difluorophenylpyridinato-N,C2)picolinato-iridium(III) (FIrpic) yielded
maximum current efficiencies up to 21 cd/A.
Introduction
been used as electron transport layer in OLEDs.5-7
However, the hole transport properties of most triazines
are not sufficient for their use as host material in OLEDs.
Recent developments of new matrix materials for phos-
phorescent organic light-emitting diodes often focus
on the synthesis of bipolar materials.8,9 In a study by
Inomata et al., donor-substituted 2,4,6-tris(diarylamino)-
1,3,5-triazines were presented as matrix material and
electron conductor for OLEDs.10 The combination of
electron-accepting triazines and electron-donating diaryl-
amine resulted in bipolar transport.11 They proved to be
a good host material for the green phosphorescent emit-
ter tris(2-phenylpyridine)iridium (Ir(ppy)3). The highest
triplet energy was presented for 2,4,6-tris(carbazolyl)-1,
3,5-triazine (ΔE(T1-S0) = 2.81 eV). An OLED with
2,4,6-tris(carbazolyl)-1,3,5-triazine as host and Ir(ppy)3
as guest have shown a maximum quantum efficiency
of 10.2%.
Since the discovery of efficient phosphorescent emitters
for organic light emitting diodes (OLED) by Forrest and
Thompson many groups face the challenge to develop
suitable new materials for phosphorescent OLEDs.1,2 In
most small molecule-based OLEDs the emitting layer
consists of a phosphorescent dye which is doped into a
host material to avoid concentration quenching and to
optimize the charge balance. High efficiencies are ob-
tained only when the energy is efficiently transferred from
the host the phosphorescent dye. For an efficient transfer
the triplet energy of the host material has to be higher
than that of the guest.3 This becomes more challenging
for host materials used in combination with blue phos-
phorescent emitters, where triplet energies larger than
2.8 eV are required. In general, the electron mobility of
many organic hosts is much lower than the hole mobility
because of the fact that they mainly consist of strong
electron donors like aromatic amines or carbazoles.4 In
contrast, electron transport materials often contain elec-
tron deficient heterocycles. Triazines, for example, are
known to be good electron conductors and therefore have
Other OLED-relevant properties like the electrochemi-
cal properties of s-triazines12 or the ab initio calculated
molecular orbitals and triplet energies13 are described for
this class of materials and show their potential as host
materials for OLEDs. In a recent work, the electric field
induced quenching on the phosphorescence intensity of a
*Corresponding author. E-mail: peter.strohriegl@uni-bayreuth.de.
(1) Baldo, M. A.; O’Brien, D. F.; You, Y.; Shoustikov, A.; Sibley, S.;
Thompson, M. E.; Forrest, S. R. Nature 1998, 395, 151–154.
(2) Yersin, H., Ed. Highly Efficient OLED’s with Phosphorescent
Materials; Wiley-VCH: Weinheim, Germany, 2008.
(8) Su, S.-J.; Sasabe, H.; Takeda, T.; Kido, J. Chem. Mater. 2008, 20,
1691–1693.
(3) Tanaka, I.; Tabata, Y.; Tokito, S. Chem. Phys. Lett. 2004, 400,
86–89.
(4) D’Andrade, B. W.; Datta, S.; Forrest, S. R.; Djurovich, P.;
Polikarpov, E.; Thompson, M. E. Org. Electron. 2005, 6, 11–20.
(5) Fink, R.; Frenz, C.; Thelakkat, M.; Schmidt, H.-W. Macro-
molecules 1997, 30, 8177–8181.
(9) Su, S.-J.; Chiba, T.; Takeda, T.; Kido, J. Adv. Mater. 2008, 20,
5951–5953.
(10) Inomata, H.; Goushi, K.; Masuko, T.; Konno, T.; Imai, T.; Sasabe,
H.; Brown, J. J.; Adachi, C. Chem. Mater. 2004, 16, 1285–1291.
(11) Son, K.; Yahiro, M.; Imai, T.; Yoshizaki, H.; Adachi, C. Chem.
Mater. 2008, 20, 4439–4446.
(6) Fink, R.; Heischkel, Y.; Thelakkat, M.; Schmidt, H.-W.; Jonda, C.;
Hueppauff, M. Chem. Mater. 1998, 10, 3620–3625.
(7) Richard, A.; Klenklera, H. A.; Tranc, A.; Popovic, D. Z.; Xu, G.
Org. Elect. 2008, 9, 285–290.
(12) Selby, T. D.; Stickley, K. R.; Blackstock, S. C. Org. Lett. 2000, 2,
171–174.
(13) Chu, T.; Ho, M.; Chen, J.; Chen, C. Chem. Phys. Lett. 2005, 415,
137–140.
r
2010 American Chemical Society
Published on Web 03/10/2010
pubs.acs.org/cm

2404 Chem. Mater., Vol. 22, No. 7, 2010
Scheme 1. Structures of the Donor-Substituted Triazines TRZ 1-7
Rothmann et al.
deep-blue triplet emitter dispersed in a donor substituted
triazine is described.14
needed for the crystallization favors the stability of
amorphous films.15,18
Another essential requirement for the successful opera-
tion of OLEDs, which is often neglected, is the ability of
the material to form stable amorphous films.15 This
property guarantees that the emitter stays uniformly
diluted in the matrix to minimize the effect of concentra-
tion quenching. In addition, the absence of grain bound-
aries, which may act as trap states, makes the use of
organic glasses as OLED materials advantageous.16,17 To
obtain organic glasses, several points have to be taken
into account. An important fact is to avoid strong inter-
molecular forces like hydrogen bonding or π-π stacking
between the molecules. Furthermore, the introduction of
bulky substituents leads to a larger intermolecular dis-
tance and a hindrance in packing and therefore an
amorphous behavior. On the other hand, the increased
distance separates the conducting units from each other
and results in decreasing charge carrier mobility. Another
possibility to design organic glasses is the synthesis of
asymmetric molecules. With this method, the number of
conformers is increased and the high amount of energy
In this paper, we describe the successful combination
of high triplet energies and amorphous behavior for
donor substituted triazines. This was achieved by the
introduction of substituted carbazoles and the successive
replacement of up to two carbazoles by phenoxy groups
(Scheme 1). This yielded asymmetric materials with glass-
transition temperatures up to 170 °C and triplet energies
up to 2.96 eV. The device performance of TRZ 2 as host
materials for the blue phosphor FIrpic is presented.
Experimental Section
Materials. All commercially available reagentswere purchased
fromAldrich and used withoutfurther purification. The synthesis
of tris[(3-phenyl-1H-benzimidazol-1-yl-2(3H)-ylidene)-1,2-phe-
nylene]-Iridium (DPBIC) and 2,8-bis(triphenylsilyl)-dibenzo-
furan (DBFSi) is described in the literature.19,20
2,4,6-Tris(methylcarbazole-90-yl)-1,3,5-triazine(TRZ 1). TRZ 1
is a statistical mixture of a triazine core with 2- and 3-methyl-
carbazole substituents. A mixture of 2-methylcarbazole (1.45 g,
8.0mmol) and3-methylcarbazole (1.45g, 8.0mmol) wasdissolved
in dry THF (100 mL) under an argon atmosphere in a two-necked
(18) Koene, B. E.; Loy, D. E.; Thompson, W. E. Chem. Mater. 1998, 10,
2235–2250.
(14) Haneder, S.; Da Como, E.; Feldmann, J.; Rothmann, M. M.;
€
Strohriegl, P.; Lennartz, C.; Molt, O.; Munster, I.; Schildknecht,
(19) Bolt, M.; Lennartz, C.; Prinz, M.; Schmidt, H.-W.; Thelakkat, M.;
Baete, M.; Neuber, C.; Kowalsky, W.; Schildknecht, C.; Johannes,
H.-H. WO Patent 2005 019373.
C.; Wagenblast, G. Adv. Funct. Mater. 2009, 19, 2416–2422.
(15) Strohriegl, P.; Grazulevicius, J. V. Adv. Mater. 2002, 14, 1439–
1452.
(20) Langer, N.; Kahle, K.; Lennartz, C.; Molt, O.; Fuchs, E.; Rudolph,
J.; Schildknecht, C.; Watanabe, S.; Wagenblast, G.; WO Patent
2009, 003898.
(16) Shirota, Y. J. Mater. Chem. 2000, 10, 1–25.
(17) Schmechel, R.; Von Seggern, H. Phys. Stat. Sol. (A) 2004, 201,
1215–1235.

Article
Chem. Mater., Vol. 22, No. 7, 2010 2405
round-bottom flask equipped with a septum inlet and three way
stopcock. n-Buthyllithium (1.6 M in hexane solution) (9.37 mL,
15 mmol) was added dropwise to the methylcarbazole solu-
tion and the mixture was stirred for 10 min. In a separate
three-necked round-bottom flask equipped with a septum inlet,
three-way stopcock, and condenser cyanuric chloride (0.83 g.
4.5 mmol) was dissolved in dry THF (60 mL) under argon. The
carbazole-lithium solution was added dropwise to the cyanuric
chloride solution using a transfer canula within 15 min. The reac-
tion mixture was refluxed for 6 h. After the solution was cooled to
room temperature, 40 mL of water were added. The product was
filtered off; washed with water, hexane, and diethyl ether; and
further prepurified by hot filtration from ethanol. Recrystalliza-
tion from a chlorobenzene/hexane mixture yielded 1.7 g (61%) of
TRZ 1 as a white crystalline solid. 1H NMR (250 MHz, d4-THF):
δ (ppm) 8.96 (m, 6H), 8.02 (m, 6H), 7.31 (m, 9H), 2.53 (s, 4.3H),
2.38 (s, 4.7H). EI-MS: m/z 618 (100, Mþ), 309 (30, M2þ).
2,4-Bis(carbazol-90-yl)-6-chloro-1,3,5-triazine (1a). Carbazole
(1.42 g, 8.5 mmol) was dissolved in dry THF (50 mL) under
argon in a two-necked, round-bottom flask equipped with a
septum inlet and three way stopcock. n-Buthyllithium (1.6 M in
hexane solution) (5.0 mL, 8.0 mmol) was added dropwise to the
methylcarbazole solution and the mixture was stirred for 10 min.
In a separate three-necked, round-bottom flask equipped with a
septum inlet, three-way stopcock, and condenser, cyanuric
chloride (0.74 g, 4.0 mmol) was dissolved in dry THF (20 mL)
in an argon atmosphere. The carbazole lithium solution was
added dropwise to the cyanuric chloride solution using a
transfer canula within 10 min. The reaction mixture was re-
fluxed for 6 h. After the solution was cooled to room tempera-
ture, 40 mL of water was added. The product was filtered off;
washed with water, hexane, and diethyl ether; and further
purified by hot filtration from ethanol, to yield 1.2 g (68%) of
1a as a white crystalline solid. Mp: 231 °C. 1H NMR (250 MHz,
CDCl3): δ (ppm) 8.92 (d, 4H), 8.06 (d, 4H), 7.56-7.40 (m, 8H).
EI-MS: m/z 445 (100, Mþ), 192 (90), 166 (100).
2,4-Bis(30-methylcarbazol-90-yl)-6-chloro-1,3,5-triazine (1b).
This compound was synthesized by a procedure similar to that
of 1a except that 3-methylcarbazole was used instead of carba-
zole. Yield: 1.5 g (78%) as a white solid. Mp: 235 °C. 1H NMR
(250 MHz, CDCl3): δ (ppm) 8.81 (d, 2H), 8.66 (d, 2H), 7.92
(d, 2H), 7.72 (s, 2H), 7.41-7.33 (m, 4H), 7.22 (d, 2H), 2.50
(s, 6H). EI-MS: m/z 473 (100, Mþ), 237 (20, M2þ), 180 (15).
2,4-Bisphenoxy-6-chloro-1,3,5-triazine (2). Cyanuric chloride
(3.68 g, 20 mmol) was dissolved in acetone (200 mL) and cooled
to 15 °C. In a separate flask, phenol (3.76 g, 40 mmol) was
reacted with sodium hydroxide (1.60 g, 40 mmol) in water
(200 mL). After the formation of the sodium phenolate was
complete, the solution was cooled to 15 °C and added dropwise
to the cyanuric chloride solution. The reaction mixture was
stirred at temperatures below 15 °C for 8 h before 200 mL of
water was added. The white precipitate was filtered and washed
with water and ethanol. The product was purified by recrystal-
lization from hexane to yield 5.8 g (85%) as white crystals. 1H
NMR (250 MHz, CDCl3): δ (ppm) 7.42-7.35 (m, 4H),
7.28-7.22 (m, 2H), 7.13 (d, 4H). EI-MS: m/z 299 (100, Mþ).
2,4-Bis(carbazol-90-yl)-6-phenoxy-1,3,5-triazine (TRZ 2). In a
two-necked, round-bottom flask, 2,4-bis(carbazol-90-yl)-6-
chloro-1,3,5-triazine (1a) (0.89 g, 2 mmol) was dissolved in
chloroform (20 mL). In a separate flask, phenol (0.23 g, 2.4
mmol) and sodium hydroxide (0.095 g, 2.4 mmol) were dissolved
in 10 mL of water at room temperature and stirred for 10 min
until the solution was added to the 1a. Tetrabutylammonium
bromide (12 mg, 0.04 mmol) and dibenzo-18-crown-6 (14 mg,
0.04 mmol) were added as phase-transfer catalysts. The mixture
is heavily stirred for 8 h at 70 °C. After cooling of the solution
to room temperature, addition of 20 mL of water, and extrac-
tion with diethyl ether, the organic phase was washed with water
and dried over magnesium sulfate. The solvent was removed
under reduced pressure and the product was purified by recrys-
tallization from a hexane/THF mixture to yield 0.40 g (40%)
of TRZ 2 as a white crystalline solid. 1H NMR (250 MHz,
CDCl3): δ (ppm) 8.74-8.70 (m, 4H), 8.04-8.00 (m, 4H), 7.60 (d,
2H), 7.44-7.36 (m, 11H). EI-MS: m/z 503 (100, Mþ), 337 (50),
166 (40).
2-(Carbazol-90-yl)-4,6-bisphenoxy-1,3,5-triazine (TRZ 3). This
compound was synthesized by a procedure similar to that of
1a except that the ratio carbazole to triazine was 1.2 to 1. Yield:
1
0.77 g (90%) as white crystalline solid. H NMR (250 MHz,
CDCl3): δ (ppm) 8.07 (d, 2H), 7.89 (d, 2H), 7.60-7.53 (m, 4H),
7.46-7.41 (m, 2H), 7.36-7.25 (m, 6H), 7.13-7.07 (m, 2H). EI-
MS: m/z 430 (100, Mþ), 337 (100), 166 (50).
2,4-Bis(carbazol-90-yl)-6-(30,50-dimethylphenoxy)-1,3,5-triazine
(TRZ 4). This compound was synthesized by a procedure similar
as described for TRZ 2. 3,5-dimethylphenol was used instead of
phenol. Yield: 0.92 g (86%) as a white crystalline solid. 1H NMR
(250 MHz, CDCl3): δ (ppm) 8.78-8.74 (m, 4H), 8.05-8.01
(m, 4H), 7.41-7.29 (m, 8H), 7.09 (s, 1H), 7.04 (s, 2H), 2.44 (s,
6H). EI-MS: m/z 531 (70, Mþ), 365 (100), 166 (50), 105 (40).
2,4-Bis(30-methylcarbazol-90-yl)-6-phenoxy-1,3,5-triazine (TRZ 5).
This compound was synthesized by a procedure similar as de-
scribed for TRZ 2. Instead of 2,4-bis(carbazol-90-yl)-6-chloro-
1,3,5-triazine (1a) 2,4-bis(30-methylcarbazol-90-yl)-6-chloro-1,3,
5-triazine (1b) was used. Yield: 0.92 g (86%) as a white crystalline
solid. 1H NMR (250 MHz, CDCl3): δ (ppm) 8.72-8.68 (m, 2H),
8.59 (d, 2H), 7.99-7.95 (m, 2H), 7.79 (s, 2H), 7.57 (d, 2H),
7.45-7.33 (m, 7H), 7.14 (d, 2H), 2.52 (s, 6H). EI-MS: m/z 531
(100, Mþ), 351 (30), 180 (30).
2,4-Bis(30-methylcarbazol-90-yl)-6-(30,50-dimethylphenoxy)-1,
3,5-triazine (TRZ 6). This compound was synthesized by a
procedure similar as described for TRZ 2. Instead 2,4-bis-
(carbazol-90-yl)-6-chloro-1,3,5-triazine (1a) 2,4-bis(30-methyl-
carbazol-90-yl)-6-chloro-1,3,5-triazine (1b) was used. 3,5-Dim-
ethylphenol was used instead of phenol. Yield: 0.86 g (81%) as a
1
white crystalline solid. H NMR (250 MHz, CDCl3): δ (ppm)
8.70-8.68 (m, 2H), 8.59 (d, 2H), 7.97-7.94 (m, 2H), 7.78
(s, 2H), 7.57 (d, 2H), 7.45-7.33 (m, 4H), 7.10 (s, 1H), 7.04 (s,
2H), 2.50 (s, 6H), 2.43 (s, 6H). EI-MS: m/z 559 (100, Mþ), 379
(50), 180 (35).
Bis(4,6-di(30-methylcarbazol-90-yl)-1,3,5-triazin-2-yl)resorcinol
(TRZ 7). This compound was synthesized by a procedure
similar to that described for TRZ 2. Instead of 2,4-bis(carbazol-
90-yl)-6-chloro-1,3,5-triazine (1a) 2,4-bis(30-methylcarbazol-90-yl)-
6-chloro-1,3,5-triazine (1b) was used. Instead of phenol, resorcinol
was used. The ratio of triazine to resorcinol to sodium hydroxide
was changed to 1:2.4:2.2 (mmol). Yield: 0.47 g (48%) as a white
1
solid. Mp: 300 °C. H NMR (250 MHz, CDCl3): δ (ppm) 8.74
(d, 4H), 8.58 (d, 4H), 7.93 (d, 4H), 7.74 (s, 4H), 7.60 (s, 1H), 7.50
(d, 2H), 7.42-7.25 (m, 9H), 7.21-7.13 (d, 4H), 2.49 (s, 12H).
NMR. EI-MS: m/z 984 (100, Mþ), 804 (20), 492 (40, M2þ),
180 (100).
Characterization. 1H NMR spectra were recorded on a Bru-
ker AC-250 spectrometer at 250 MHz. All NMR data given are
given as chemical shifts δ (ppm) downfield from Si(CH3)4. MS
spectra were obtained on a Finnigan Mat 8500, MAT 112 S
Varian machine using EI-ionization. Melting points and
glass-transition temperature determination was carried out
on a Perkin-Elmer Diamond DSC at a heating/cooling rate of

2406 Chem. Mater., Vol. 22, No. 7, 2010
Rothmann et al.
10 K/min under a nitrogen atmosphere. Thermographimetric
analysis (TGA) was performed on a Mettler Toledo TGA/
SDTA851e machine at a heating rate of 10 K/min in nitrogen
atmosphere.
properties of these triazines were still poor. We found
that a stoichiometric mixture of 2-methylcarbazole and
3-methylcarbazole (1:1) as substituents was the most
effective way to obtain a less-crystalline material. TRZ
1 is synthesized by nucleophilic substitution of cyanuric
chloride with a 1:1 mixture of 2- and 3-methyl substituted
carbazoles (see Scheme 2). The H NMR spectrum of
TRZ 1 confirmed that the carbazoles are incorporated in
an almost equimolar ratio.
The synthesis of the unsymmetrical triazines TRZ 2 to
TRZ 7 followed a different route. To obtain these com-
pounds, at least two steps are required. For all materials
except for TRZ 3, the first synthetic step was a substitu-
tion of cyanuric chloride with carbazole or 3-methylcar-
bazole. The number of carbazole substituents was con-
trolled by the ratio of carbazole and triazine. This is
possible because of the stepwise deactivation of the
chlorine atoms at the triazine core after substitution with
an electron donating carbazole unit. Therefore, the reac-
tion can be efficiently stopped at this step and yields of
68% (1a) and 78% (1b) were obtained. In the second step,
1a and 1b were reacted with phenol and 3,5-dimethylphe-
nol to yield triazines TRZ 2 and TRZ 5 as well as TRZ 4,
and TRZ 6, respectively. In contrast to that, phenoxy
substitutions mainly follow a temperature control.28 The
synthesis of intermediate 2 was carried out at tempera-
tures below 15 °C to suppress the 3-fold substitution and
yielded 85%. The substitution of 2 with carbazole yielded
the triazine TRZ 3. For the synthesis of triazine TRZ 7,
the diol resorcinol was reacted with 1b.
UV-vis absorption spectra were recorded on a Hitachi
U-3000 spectrometer. Photoluminescence spectra were obtained
using a Shimadzu Spectrofluorophotometer RF-5301PC. UVA-
SOL solvents purchased from Merck were used to prepare
solutions and films for optical characterization. The phosphor-
escence spectra were obtained with a time gated PL setup
described elsewhere.14 In brief, thin film samples were mounted
in a liquid helium coldfinger cryostat and excited by laser pulses
(500 ps) at 337 nm. The detection of phosphorescence was
accomplished by a gated microchannel-plate (MCP) with a
time-gated detection window delayed with respect to the photo-
exciting pulse. For the experiments presented here, the detection
started 500 ns after the pulse and for a time interval of 1 ms.
OLED Fabrication. The organic layers were deposited by
thermal evaporation in high vacuum (<1 ꢀ 10-6 mbar) onto
indium-tin-oxide (ITO, 10 ohm/square) precoated glass sub-
strates. Prior to use the ITO glass was degreased using organic
solvents and cleaned using an UV-ozone oven for 30 min. The
organic layers and the metal cathode were evaporated without
breaking the vacuum.
1
Computational Calculations. The transport levels of the ma-
terials used were determined via density functional calculations.
For the ionization potential and the electron affinity, first the
geometry of the neutral as well as the charged states was
optimized using the BP86-functional21,22 in combination with
a split-valence basis set (SV(P)) including polarization functions
on all heavy atoms.23 For iridium, an effective core potential was
employed.24 For the energetics, single-point calculations at the
optimized geometries using the same functional in combination
with a valence triple-ζ basis set (TZVP)25 were conducted. To
account for dielectric solid-state effects, we used a UPS/IEPS-
calibrated version of the conductor-like screening model
(COSMO)26 in conjunction with the single-point calculations.
All calculations were carried out with the turbomole program
package.27
Thermal Characterization. The thermal properties of
the novel triazines were investigated by differential scan-
ning calorimetry (DSC), and their thermal stability was
examined by thermogravimetric analysis (TGA). In both
experiments, the heating and cooling rate was 10 K/min.
The data are shown in Table 1. The first and second
heating as well as the first cooling curves of TRZ 1, TRZ
4, and TRZ 6 are shown in Figure 1.
Results and Discussion
All triazines exhibit a melting peak during the first
DSC heating cycle. The melting temperatures range from
147 °C for TZR 5 to 335 °C for TRZ 1 (see Figure 1). TRZ
1, TRZ 2 and TRZ 3 crystallize upon cooling in the DSC.
TRZ 4 forms a glass upon cooling and a glass transition
can be observed in the subsequent heating cycle at 83 °C.
On further heating, TRZ 4 crystallizes at 127 °C and
exhibits a melting at 210 °C (see Figure 1). TRZ 5, TRZ 6
(see Figure 1), and TRZ 7 also form glasses on cooling. In
contrast to TRZ 4, these triazines stay amorphous in the
following heating and cooling cycles. The glass-transition
temperatures of TRZ 5 (85 °C) and TRZ 6 (90 °C) are
slightly higher than that of TRZ 4. The higher molecular
weight of the dimeric compound TRZ 7 results in a much
higher Tg of 170 °C. The thermal behavior can be re-
produced in multiple cycles. The glass transition of TRZ 2
was determined in a separate experiment. The material
was heated above its melting temperature and afterward
Synthesis. Previously reported 2,4,6-triscarbazolyl-
1,3,5-triazine10 is highly crystalline with a melting point
of 465 °C. Several attempts have been made to reduce the
crystallinity of triscarbazolyl-triazines. The substitution
of the s-triazine core with three 2-methylcarbazole units
or three 3-methylcarbazoles still yielded materials with a
high tendency to crystallize. Furthermore, we synthesized
unsymmetrically substituted triscarbazolyl triazines. In a
two-step synthesis, we were able to prepare various
compounds with either two methylcarbazoles and one
carbazole unit or three different methylcarbazole units.
This strategy resulted in materials with slightly decreased
crystallization tendencies. However, the glass formation
(21) Perdew, J. P. Phys. Rev. B 1986, 3319, 8822.
(22) Becke, A. D. Phys. Rev. A 1988, 36, 3098.
€
(23) Schafer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 9, 2571.
(24) Andrae, D.; Haeussermann, U.; Dolg, M.; Stoll, H.; Preuss, H.
Theor. Chim. Acta 1990, 77, 123.
€
(25) Schafer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 1994, 100, 5829.
(26) Klamt, A. J. Phys. Chem. 1995, 99, 2224.
€
€
€
(27) Ahlrichs, R.; Bar, M.; Haser, M.; Horn, H.; Colmel, C. Chem.
(28) Schaefer, F. C.; Thurston, J. T.; Dudley, J. R. J. Am. Chem. Soc.
1951, 73, 2990–2992.
Phys. Lett. 1989, 162, 165.

Article
Chem. Mater., Vol. 22, No. 7, 2010 2407
Scheme 2. Synthesis of Donor-Substituted Triazines TRZ 1-7
Table 1. Thermal Properties of Triazines TRZ 1-TRZ 7a
these temperature values for all materials. This high
thermal stability enables thermal evaporation without
decomposition. It was possible to yield amorphous films
by spin coating or thermal evaporation from all triazine
host materials. The films of TRZ 1 and TRZ 3 tended to
crystallize after several hours to days. All other films were
stable for at least 3 months (period of observation).
Optical Properties. The absorption, fluorescence, and
phosphorescence spectra of TRZ 1-TRZ 3 films (thick-
ness: 150 nm) on quartz are shown in Figure 2. The optical
data for TRZ 1-3 are summarized in Table 2. The spectra
show typical examples of triazines with three, two, and
one carbazole group. The spectra of TRZ 4-TRZ 7 are
very similar to those of TRZ 2 with a little bathochromic
shift due to the inductive effect of the methyl groups
attached to the carbazole. By comparing the absorption
spectra of three triazines, it can be observed that the
lowest energetic absorption maximum is shifted from 333
to 320 nm with an increasing amount of phenoxy substit-
uents. From the absorption edge of these spectra, the
optical bandgap can be estimated. TRZ 1 has the lowest
bandgap of 3.54 eV. TRZ 2 and TRZ 3 have bandgap
values of 3.61 and 3.75 eV, respectively. In the nongated
room-temperature PL spectra, we observe the fluor-
escence of the compounds and the bathochromic shift is
product
Tg (°C)
Tm (°C)
Tc (°C)
Td (°C)
TRZ 1
TRZ 2
TRZ 3
TRZ 4
TRZ 5
TRZ 6
TRZ 7
335
219
248
295
135
186
127c
372
356
307
310
364
386
399
80b
83
85
90
170
210
147d
171d
301d
aTg, glass-transition temperatures; Tm, melting temperatures; Tc,
crystallization temperatures; and Td, onset temperatures of decomposi-
tion or evaporation at ambient pressure. b Observed only after very fast
cooling of the melt. c Observed only upon heating. d Observed only in the
first heating scan.
cooled very fast using liquid nitrogen to yield a super-
cooled melt. On subsequent heating in the DSC, a Tg was
observed at 80 °C followed by a crystallization peak at
121 °C and melting peak at 219 °C. However, it was not
possible to observe a glass transition of TRZ 1 and TRZ 3
using this technique. These experiments demonstrate that
it is possible to get stable glasses from these triazine
compounds. All compounds exhibit a high thermal sta-
bility in the TGA experiment. The onset temperatures
range from 307 to 399 °C for TRZ 3 and TRZ 7,
respectively (see Table 1). The evaporation temperatures
in ultrahigh vacuum were found to be at least 50 °C below

2408 Chem. Mater., Vol. 22, No. 7, 2010
Rothmann et al.
Figure 1. First heating cycle (top line), second heating cycle (middle line), and first cooling cycle (bottom line) of TRZ 1, TRZ 4, and TRZ 6 (left to right).
Table 2. Photoluminescence Properties of Triazines TRZ 1-TRZ 3a
ΔE(S0-S1)
(eV)
λfluor
(nm)
λphos
(nm)
ΔE(T1-S0)
(eV)
product
λabs (nm)
TRZ 1 285, 333
TRZ 2 281, 330
TRZ 3 280, 308,
320
3.54
3.61
3.75
385
433, 465
2.86
2.95
2.96
372, 389 420, 450
351 419, 450
a λabs, maxima of the absorption; ΔE(S0-S1), singlet bandgap, calcu-
lated form the edge of the absorption spectra; λfluor, maxima of the
fluorescence; λphos, maxima of the phosphorescence; ΔE(T1-S0), triplet
energy, calculated from the 0-0 phosphorescence transition.
and in the presence of the long-lived phosphorescence.
The triplet energy was obtained from the first vibronic
transition of the red-shifted phosphorescence spectrum.
All spectra exhibited a vibronic progression up to the 0-2
peak. TRZ 1 shows a 0-0 phosphorescence transition at
433 nm (2.86 eV) and a maximum at 465 nm. Values of
420 nm (2.95 eV) and 419 nm (2.96 eV) for TRZ 2 and
TRZ 3 are determined for the respective 0-0 transitions.
In contrast to the fluorescence, the difference of the first
phosphorescence maximum of TRZ 2 and TRZ 3 is only
1 nm. These maxima are centered at 450 nm and corre-
spond to the 0-1 transition. The weak signals between
320 and 400 nm may result from delayed fluorescence of
the host materials as suggested by the good energy over-
lap with the prompt fluorescence spectra (dashed lines).
From the high triplet energy values, it can be estimated
that efficient exothermic energy transfer from all the
presented TRZ host to blue phosphors such as FIrpic
(2.65 eV) is possible.
Phosphorescent Organic Light-Emitting Diodes. Two
OLEDs with the configurations shown in Figure 3
were fabricated to demonstrate the potential of donor-
substiuted triazines as host material for blue phosphores-
cent emitters. For this purpose, TRZ 2 was chosen as
representative material for this class. On top of indium-
tin-oxide (ITO), poly(thiophene-3-[2-(2-methoxyethoxy)-
ethoxy]-2,5-diyl) was used as hole-injection layer followed
by 35 nm of DPBIC p-doped with molybdenum(VI)
oxide as hole-transporting layer. An additional 10 nm
thick layer of DPBIC followed as exciton and electron
blocker. The20nm thick emission layer consistedofTRZ 2
doped with 8% (device 1) or 16% (device 2) FIrpic. In
device 1, TRZ 2 was used as hole-blocking layer (5 nm)
followed by 20 nm 2,9-dimethyl-4,7-diphenyl-1,10-phe-
nanthroline (BCP) as electron transporting material. In
device 2, the hole-blocking and electron-conduction layers
Figure 2. Normalized absorption (left solid line), fluorescence (dotted
line), and phosphorescence (right solid line) spectra of neat films of TRZ
1, TRZ 2, and TRZ 3 (top to bottom).
even more pronounced. The fluorescence of TRZ 1 is
relatively broad with a maximum at 385 nm. The spec-
trum of TRZ 2 has two sharp peaks with almost the same
height at 372 and 389 nm and a broad red-shifted
shoulder. The fluorescence of TRZ 3 is narrower with a
maximum at 351 nm. The hypsochromic shift can be
assigned to the electron withdrawing effect of the triazine
on the single carbazole. In TRZ 3, the effect is very
pronounced because only one carbazole unit is present.
The hypsochromic shift becomes smaller in TRZ 2 and
TRZ 1 with two and three carbazole units connected to
the central triazine.
For the application as host material in OLEDs, the
triplet level of the host must be higher compared to the
guest. The triplet energy of selected triazines was deter-
mined from neat film time-gated phosphorescence spec-
tra of neat films at T = 5 K. In this experiment, the
detection of the phosphorescence spectrum is achieved by
recording the PL several hundred nanoseconds after the
laser pulse, i.e., after the prompt fluorescence has decayed

Article
Chem. Mater., Vol. 22, No. 7, 2010 2409
Figure 3. Top: Electroluminescent spectra of device 1 and device 2 at 10 mA/cm2. Bottom: Energy level diagram of device 1 and device 2; HOMO and
LUMO levels of the different materials are indicated. The dotted lines represent the levels of the emitter FIrpic. The HOMO and LUMO levels were
extracted from density functional theory (DFT) calculations.
consisted of 5 nm DBFSi and 40 nm BCP. The devices
were finalized by deposition of LiF (0.7 nm)/Al (100 nm)
cathode.
In the literature, the bipolar transport characteristics of
a carbazol substituted triazine were reported.10 Therefore
we consider it very important that exciton blocking layers
are used on both sides of the emission layer. In our setup,
DPBIC is used as exciton and electron blocker on the hole
conduction side. The blocker on the electron conduction
side was varied. In device 1, the host material TRZ 2 itself
was used. DBFSi, which was employed in device 2, acts
both as effective hole and exciton blocker.
Figure 3 considers the electroluminescence spectra of
both devices at a current density of 10 mA/cm2. Both
spectra show a pure emission of FIrpic with a maximum
at 473 nm and CIE coordinates of x = 0.17 and y = 0.32.
Figure 4(top) displays the current density-voltage-
luminance characteristics of the devices. The device with
TRZ 2 as hole and excition blocking material (device 1)
exhibted a 0.8 V lower turn-on voltage. This voltage in-
crease can be attributed to the application of DBFSi as
blocking layer which exhibits a higher LUMO level than
TRZ 2 (see Figure 3). This results in an additional injection
barrier. On the other side the current density and the
Figure 4. Top:Luminance-voltage(solid symbols) and current-voltage
(open symbols) characteristics of device 1 (circles) and device 2 (triangles).
luminance curves were steeper in device 2. Maximum
luminances of 6900 and 1500 cd/m2 were reached at 11 V
for device 2 and device 1, respectively. The higher value of
device 2 was attributed to the more effective exciton
blocking of DBFSi. Figure 4(bottom) shows the external
quantum efficiency (EQE) and current efficiency plotted
versus the voltage. Here, the efficiency of the different
blockers is even more pronounced. Device 2 exhibts a
Bottom: External quantum efficiency-current density (solid symbols)
and luminous efficiency-current density plots (open symbols) of device
1 (circles) and device 2 (triangles).
maximum EQE value of 10.2% at 22 cd/m2. The maxi-
mum luminous and power efficiencies are 20.4 cd/A and
13.2 lm/W, respectively. At a luminesence of 100 cd/m2, an
EQE of 9.5% and a luminous efficiency of 19.2 cd/A were
reached. The efficiencies of device 1 were significantly

2410 Chem. Mater., Vol. 22, No. 7, 2010
Rothmann et al.
lower. Maximum external quantum, luminous, and power
efficiencies of 6.7%, 12.6 cd/A, and 8.7 lm/W were reached
at5.1, 5.1, and 4.4 V, respectively. Thesemuchlower values
highlight the importance of a suitbale hole/exciton block-
ing layer in this device setup. These results demonstrate the
potential of donor-substituted triazines as host materials
for blue phosphorescent OLEDs.
morphological stability. The triplet energies of the tria-
zines were in the range from 2.86 to 2.96 eV. The potential
of this class of materials as hosts for blue phosphors has
been tested. A maximum brightness of 6900 cd/m2 and a
maximum external quantum efficiency of 10.2% were
reached using the blue emitter FIrpic. These first results
show that the donor substituted of triazines hold great
potential as host materials for blue OLEDs.
Conclusion
Acknowledgment. We thank C. Bonsignore for his support
during the preparation and characterization of the light-
emitting diodes. Our special thanks go to I. Bauer for
supporting us during synthesis and characterization of the
In this paper, we presented a series of new matrix
materials based on the concept of donor-susbtituted
triazines with three, two, and one carbazole unit. The
thermal properties, especially the glass formation, were
controlled by different substitution patterns and amor-
phous host materials with glass-transition temperatures
up to 170 °C are obtained. This guarantees a long-term
€
host materials. We also thank Dr. Ingo Munster, Dr. Evelyn
Fuchs, and Dr. Nicolle Langer for scientific discussions. We
are grateful to the BMBF for financial support through the
project OPAL 2008.