Engineering blue fluorescent bulk emitters for OLEDs: Triplet harvesting by green phosphors

Article abstract of DOI:10.1021/cm500602y

Triplet harvesting in organic light-emitting diodes (OLEDs) from a blue fluorescent to a green phosphorescent emitter is important for the development of highly efficient white OLEDs for lighting applications. Here, we report new blue fluorescent bulk emitters with high triplet energies for triplet harvesting in OLEDs. Based on the chemical structure of the highly efficient blue emitter N,N′-di-1-naphthalenyl-N,N′-diphenyl-[1,1′:4′,1″: 4″,1′-quaterphenyl]-4,4′ diamine (4P-NPD), two new quarterphenylene compounds are designed and synthesized. Their experimentally obtained ionization potentials, singlet, and triplet levels are compared to quantum chemical calculations. We show that the incorporation of repulsive methyl groups leads to a twist of the outer phenyl rings of the molecule and increases the triplet level. Finally, triplet harvesting OLEDs comprising these two new compounds as bulk emitters are electrically and optically characterized. We demonstrate that both materials allow triplet harvesting by the green phosphorescent emitter iridium(III)bis(2-phenylpyridinato-N,C2′) acetylacetonate (Ir(ppy)₂(acac)).

Full text of DOI:10.1021/cm500602y

Article
pubs.acs.org/cm
Engineering Blue Fluorescent Bulk Emitters for OLEDs: Triplet
Harvesting by Green Phosphors
Simone Hofmann,*,† Markus Hummert,†,∥ Reinhard Scholz,† Regina Luschtinetz,‡ Caroline Murawski,†
Paul-Anton Will,† Susanne I. Hintschich,†,⊥ Jorg Alex,† Vygintas Jankus,§,∥ Andrew P. Monkman,§
̈
Bjorn Lussem,†,# Karl Leo,† and Malte C. Gather†,$
̈
̈
†Institut fur Angewandte Photophysik, Technische Universitat Dresden, George-Bahr-Straße 1, 01069 Dresden, Germany
‡Institut fur Physikalische Chemie und Elektrochemie, Technische Universitat Dresden, Mommsenstraße 13, 01062 Dresden,
̈
̈
̈
̈
̈
Germany
§Physics Department, University of Durham, South Road, Durham DH1 3LE, United Kingdom
ABSTRACT: Triplet harvesting in organic light-emitting
diodes (OLEDs) from a blue fluorescent to a green
phosphorescent emitter is important for the development of
highly efficient white OLEDs for lighting applications. Here,
we report new blue fluorescent bulk emitters with high triplet
energies for triplet harvesting in OLEDs. Based on the
chemical structure of the highly efficient blue emitter N,N′-di-
1-naphthalenyl-N,N′-diphenyl-[1,1′:4′,1″:4″,1‴-quaterphen-
yl]-4,4‴ diamine (4P-NPD), two new quarterphenylene
compounds are designed and synthesized. Their experimen-
tally obtained ionization potentials, singlet, and triplet levels
are compared to quantum chemical calculations. We show that the incorporation of repulsive methyl groups leads to a twist of
the outer phenyl rings of the molecule and increases the triplet level. Finally, triplet harvesting OLEDs comprising these two new
compounds as bulk emitters are electrically and optically characterized. We demonstrate that both materials allow triplet
harvesting by the green phosphorescent emitter iridium(III)bis(2-phenylpyridinato-N,C2′)acetylacetonate (Ir(ppy)2(acac)).
fluorescent emitters while additionally using the radiative triplet
state of the phosphor.9,10 Here, singlets decay radiatively on the
blue fluorescent emitter, while triplets are harvested by the
phosphorescent emitter and contribute to the emission at
longer wavelengths. This allows designing highly efficient white
OLEDs.11,12 However, TH by green phosphorescent emitters is
particularly challenging since the triplet level of the blue
fluorescent emitters needs to be resonant with or above the
level of the green phosphor, a requirement that blue fluorescent
emitters have not met so far.
In the following, we summarize some of the previous efforts
on TH by green emitters. For simplicity, the triplet energy of
the different materials involved is approximated by their peak
emission wavelength λmax, i.e. T1→S0 = (2πℏc)/λmax. Sun et al.9
and Kondakova et al.13 demonstrated TH using the green
phosphorescent emitter fac-tris(2-phenylpyridine) iridium (Ir-
(ppy)3, T1→S0 = 2.4 eV). In their work, exciton formation
occurs on the matrix material 4,4′-bis(N-carbazolyl)biphenyl
(CBP, T1→S0 = 2.61 eV13), where the blue emitter is
embedded as a dopant. However, the matrix material CBP
exhibits a large HOMO−LUMO energy gap (HOMO <
1. INTRODUCTION
Organic light-emitting diodes (OLEDs) are promising
candidates for modern lighting and display applications.1,2
Due to spin statistics, a singlet to triplet ratio of 1:3 is obtained
by the recombination of holes and electrons within the OLED
structure.3 Consequently, in fluorescent emitter systems that
emit light only from the singlet state, 75% of the injected charge
carriers are lost. Therefore, major research efforts are dedicated
to phosphorescent emitters which have large intersystem
crossing rates and emit light efficiently from the triplet state.4
Although green and red phosphorescent emitters can reach
high efficiencies and long lifetimes, the development of a stable
blue phosphorescent emitter remains challenging.5,6
Furthermore, most blue phosphorescent emitters achieve
only sky-blue color coordinates, whereas white light sources
need a deep blue component to achieve a high color rendering
and meet the requirements for indoor applications. Recently,
fluorescent emitters gained much interest due to the efficient
upconversion of triplets into singlets, either via thermally
activated delayed fluorescence or via triplet fusion.7,8 With
these promising approaches, one might fully avoid the use of
phosphorescent materials one day.
The concept of triplet harvesting (TH) from a blue
fluorescent emitter by a green or red phosphorescent material
benefits from the deep blue emission and stability of the blue
Received: June 11, 2013
Revised: March 7, 2014
Published: March 7, 2014
© 2014 American Chemical Society
2414
dx.doi.org/10.1021/cm500602y | Chem. Mater. 2014, 26, 2414−2426

Chemistry of Materials
Article
−5.68 eV13, LUMO = −1.96 eV13) resulting in high driving
voltages and impeding the achievement of high luminous
efficacies.
Carbazole-based compounds are promising candidates as
deep blue bulk emitters with sufficiently high triplet energies to
provide TH by Ir(ppy)3.14,15 Yang et al.16 showed hybrid white
OLEDs with EQEs exceeding 10% (15 lm W−1) using the
carbazole/benzimidazole-based compound Cz-2pbb (T1→S0 =
2.46 eV). Recently, Zheng et al.17 reported on 7-(diphenyla-
mino)-4-methoxycoumarin (DPMC) and di[4-(4-diphenylami-
nophenyl) phenyl] sulfone (DAPSF) as ambipolar blue bulk
emitters with triplet energies above 2.45 eV. They demon-
strated white OLEDs where green and red emission is obtained
by TH, leading to efficiencies comparable to fully phosphor-
escent OLEDs (max. 22% EQE, 27 lm W−1 at 1000 cd m−2).
The successful implementation of TH in white OLEDs
underpins the importance of developing new fluorescent bulk
emitters.
Schwartz et al.18 and Rosenow et al.12 showed highly efficient
white OLEDs with more than 30 lm W−1 at a luminance of
1000 cd m−2 using doped transport layers19 and TH from the
primarily hole transporting blue bulk emitter N,N′-di-1-
naphthalenyl-N,N′-diphenyl-[1,1′:4′,1″:4″,1‴-quaterphenyl]-
4,4‴ diamine (4P-NPD, T1→S0 = 2.30 eV13).
Figure 1a illustrates the principle of TH using 4P-NPD: Due
to the preferred hole transport, excitons are generated next to
the hole blocking layer (HBL). Singlets have short lifetimes
(ns) and decay radiatively, while triplets have longer lifetimes
(ms) and can therefore diffuse toward a phosphorescent
emitter doped into 4P-NPD at an appropriate distance away
from the generation zone. At the phosphorescent emitter, 4P-
NPD triplets are harvested, leading to efficient phosphorescent
emission. At the same time, singlet transfer to the phosphor is
avoided due to the shorter diffusion length of singlet excitons in
4P-NPD (4−5 nm20) with respect to triplets (11−15 nm21). In
this manner, TH by the red phosphor iridium(III)bis(2-
methyldibenzo-[f,h]quinoxaline) (acetylacetonate) (Ir-
(MDQ)2(acac)) or by the yellow phosphor bis(2-(9,9-
dihexylfluorenyl)-1-pyridine)(acetylacetonate) iridium(III)
(Ir(dhfpy)2(acac)) was demonstrated.10−12,22
Figure 1. (a) Principle of triplet harvesting using the blue fluorescent
bulk emitter 4P-NPD. The high triplet energy of the green
phosphorescent impedes triplet harvesting from the blue fluorescent
emitter. (b) Emission spectra of TH OLEDs with 4P-NPD and a red
or green phosphorescent emitter. For the red emitter a strong spectral
contribution is observed, whereas the device with the green emitter
shows no emission from the green phosphor. The spectra are
normalized with respect to the 4P-NPD emission peak and are
vertically displaced for clarity.
However, TH from 4P-NPD to a green emitter cannot be
realized due to the low triplet level of 4P-NPD. This is
illustrated in Figure 1b which shows the emission spectrum of
two 4P-NPD based TH OLEDs comprising either the red
emitter Ir(MDQ)2(acac) or the green emitter Ir(ppy)2(acac).
(Details on the OLED structure are given in the experimental
section.) For the device comprising Ir(MDQ)2(acac), strong
red emission is observed in addition to the blue emission of 4P-
NPD, which is clear evidence for efficient TH. The device with
the green phosphor shows pure 4P-NPD emission without any
spectral contribution from the green emitter, demonstrating
that TH from 4P-NPD to the green emitter Ir(ppy)2(acac)
cannot be realized. To circumvent this problem, Schwartz et
al.18 and Rosenow et al.12 developed complex stack designs
comprising interlayers or stacked (tandem) OLEDs, both of
which require a direct recombination of holes and electrons on
a green phosphorescent emitter to achieve white light.
However, raising the triplet energy of 4P-NPD would enable
a much simpler pathway to white TH OLEDs.
(naphthalen-1-yl)-N4,N4‴-diphenyl-[1,1′:4′,1″:4″,1‴-quater-
phenyl]-4,4‴-diamine (8M-4P-NPD) and N4,N4‴-bis(9,9-
dimethyl-9H-fluoren-2-yl)-2,2′,2‴,3″,5″,6,6′,6‴-octamethyl-
N4,N4‴-diphenyl-[1,1′:4′,1″:4″,1‴-quaterphenyl]-4,4‴-dia-
mine (8M-4P-FPD), are developed and characterized in terms
of crystal structure and energy levels. For both new materials,
we demonstrate OLEDs with TH from the blue bulk emitter to
the green phosphorescent emitter iridium(III)bis(2-phenyl-
pyridinato-N,C2′) acetylacetonate (Ir(ppy)2(acac)) which has
recently been found to provide superior electroluminescence
performance in comparison to the commonly used Ir(ppy)3.23
2. EMITTER DESIGN
Our strategy to increase the 4P-NPD triplet level is the
modification of the central quarterphenyl core of 4P-NPD in
such a way that the dihedral angles between adjacent phenyl
rings are increased.
In this paper, we investigate ways to increase the triplet level
of 4P-NPD by small changes of the molecular structure. Based
on these considerations, two deep blue fluorescent bulk
emitters, 2,2′,2‴,3″,5″,6,6′,6‴-octamethyl-N4,N4‴-di-
Figure 2a shows the molecular structure of 4P-NPD and the
calculated ground state and excited state energy of the molecule
as a function of twisting angle φ between the central and the
peripheral phenyl ring of the quarterphenyl chain. The
2415
dx.doi.org/10.1021/cm500602y | Chem. Mater. 2014, 26, 2414−2426

Chemistry of Materials
Article
This finding is further corroborated by a geometric
optimization of the 4P-NPD structure in the triplet
configuration T1, which shows a substantial planarization of
the quarterphenyl chain with reduced twisting angles α = 4°
and φ = 16°. As shown in Figure 3, the planarization in the
Figure 3. Frontier orbitals of 4P-NPD: HOMO (bottom) and LUMO
(top), in the S0 ground state geometry (left) and in the T1 relaxed
excited geometry (right). The planarization in the T1 geometry allows
both frontier orbitals to localize more strongly on the quarterphenyl
chain, compare Table 1 for the respective dihedral angles between the
phenyl rings.
lowest triplet configuration shrinks the frontier orbitals toward
the quarterphenyl core region, reducing in turn their
delocalization over the terminal naphthyl and phenyl groups.
This stronger localization of the frontier orbitals increases the
electron−hole interaction in the relaxed T1 geometry and
contributes to a further red shift of the triplet emission. A
similar shrinking of the LUMO has been reported for the
relaxed excited singlet geometry of another, smaller compound
with a biphenyl core, the N,N′-diphenyl-N,N′-bis(3-methyl-
phenyl)-(1,1′-biphenyl)-4,4′-diamine (TPD).24
The splitting between the S1 singlet excitation and the T1
triplet excitation of 4P-NPD depends only weakly on the
dihedral angle φ: it increases monotonously from 0.63 eV at
φ = 0° toward 0.74 eV at φ = 90°.
The most important result of the calculations presented in
Figure 2a is that an increase of the twisting angles φ beyond
35° (the angle in the optimized ground state geometry of 4P-
NPD) raises the lowest triplet excitation energy. For φ = 90°
the triplet excitation energy increases by about 0.03 eV with
respect to the transition energy at φ = 35°. This indicates that
compounds with large dihedral angles φ could show triplet
energies that are high enough to allow for efficient transfer to
the green phosphorescent emitter Ir(ppy)2(acac). In practice,
the twisting angle φ can be increased by introducing methyl
groups to the central and peripheral phenyl rings of the
quarterthphenyl chain, as shown in Figure 2b.
Figure 2. (a) Calculated potential surface energies and transition
energies of 4P-NPD as a function of the dihedral angle φ between the
phenyl rings P1 and P2 and between P3 and P4. The dihedral angle α
in the central biphenyl chain P2−P3 is kept at the ground state value
of α = 36°. Black: potential S0, light green: triplet excitation energy
T1←S0, light blue: singlet excitation energy S1←S0, dark green: excited
triplet potential T1, dark blue: excited singlet potential S1. (b)
Chemical structure of 4P-NPD, 8M-4P-NPD, and 8M-4P-FPD.
transition energy for excitation to the lowest excited singlet
level (S1←S0) and to the lowest triplet level (T1←S0) are also
shown. The dihedral angle in the central biphenyl group was
kept fixed at its optimized ground state value of α = 36° for this
calculation. See Section 8 at the end of this paper for details on
the quantum chemical calculation.
The ground state energy of 4P-NPD is lowest for a twisting
angle of φ = 35°, indicating that this is the most likely
configuration of the molecule. For larger twisting angles, the
energy of the T1 and the S1 increase significantly. Both
transition energies (S1←S0 and T1←S0) grow monotonously
with φ so that the S1 and T1 states have potential surfaces with
a minimum at a slightly lower value of φ with respect to the
ground state potential S0.
For all three compounds shown in Figure 2b, Table 1
summarizes the geometric properties in the S0 ground state
geometry and in the relaxed T1 triplet geometry, respectively.
Table 1. Summary of Dihedral Angles α and φ (in Degrees)
of the Investigated Compounds in the S0 and T1 State,
Calculated at the B3LYP/6-31G(d) Level
4P-NPD
8M-4P-NPD
8M-4P-FPD
S0, α
S0, φ
T1, α
T1, φ
36
35
4
37
89
38
89
38
89
38
89
16
2416
dx.doi.org/10.1021/cm500602y | Chem. Mater. 2014, 26, 2414−2426

Chemistry of Materials
Article
As expected, the methyl groups in 8M-4P-NPD and in 8M-4P-
FPD introduce large dihedral angles between the central and
the peripheral phenyl rings along the central quarterphenyl
chain. As described above, 4P-NPD planarizes significantly in
the relaxed T1 geometry. However, for the compounds
containing methyl groups, the dihedral angles in the T1
configuration remain at values close to the electronic ground
state S0.
Table 2. Cyclic Voltammetry (CV) Oxidation Potential in
Dichloromethane, Derived Ionization Potential (IP) as
Determined from the Fc/Fc+ Reference (eV); DFT Values
for Ionization Potential IP (eV) and HOMO Level (eV);
Molecular Radii rion Estimated Assuming a Density of
1.11 g/cm3 (Corresponding to Crystalline 8M-4P-FPD);
Polarization Shift of IP Calculated for CH2Cl2 (with
ε = 8.93); and DFT Values Corrected for the Polarization
Shift, IP(B3LYP)-P+, in eV
Detailed computational studies of 8M-4P-NPD indicate that
the steric hindrance between the methyl groups enforces
twisting angles of φ = 89° in the electronic ground state S0,
with the central dihedral angle increasing only slightly to
α = 37°. The large φ leads to a spectroscopic decoupling of the
different subgroups: HOMO-1, HOMO, LUMO, and
LUMO+1 delocalize around the two amine groups but avoid
the central biphenyl, whereas HOMO−2 and LUMO+2
localize on the biphenyl core. Enforcing C2 symmetry for
8M-4P-NPD, the orbitals HOMO and HOMO-1 differ only in
a sign change between the contributions around both amine
groups, with a splitting lower than 3 meV. Similar arguments
apply to the nearly degenerate LUMO and LUMO+1. As will
be discussed elsewhere in more detail, this allows for a
symmetry breaking from C2 to C1 and a subsequent localization
of HOMO and LUMO on the same terminal group of the
molecule.25 This behavior is in sharp contrast to 4P-NPD
where the relaxed triplet geometry still enforces a delocalization
of the frontier orbitals over the entire molecule, with some
shrinking of both HOMO and LUMO toward the quarter-
phenyl chain.
The triplet energy can be enhanced further by adding
fluorenyl side groups to the nitrogen atoms (8M-4P-FPD,
Figure 2b). As expected, the calculated singlet and triplet
excitation energies of this compound remain above the
respective values for 8M-4P-NPD, compare Table 4.
Furthermore, these additional fluorenyl groups are expected
to result in a high fluorescence yield of the material.26 Our
finding that the increase in stiffness of the molecular structure
raises the triplet energy agrees with previous studies of other
classes of molecules.15,27
4P-NPD
8M-4P-NPD
8M-4P-FPD
CV in CH2Cl2
IP in CH2Cl2
IP (B3LYP)
HOMO (B3LYP)
rion (A)
0.88
5.37
0.82
5.31
0.70
5.19
5.74
5.74
5.60
−4.87
6.41
−4.92
6.72
−4.78
7.05
P+ (eV)
1.00
0.95
0.91
IP(B3LYP)-P+
4.74
4.79
4.69
shift for molecules in the size range investigated here, and ref
30 for investigations of a modified vacuum level relying on UPS
of molecular films. Despite possible difficulties with electron
injection, we consider 4P-NPD compounds with strongly
twisted phenyl rings as a promising route to increase the triplet
energy and to provide TH by green phosphorescent emitters in
OLEDs.
3. SYNTHESIS
The two 4P-NPD type materials with incorporated methyl
groups and different side groups were synthesized according to
the route shown in Figure 4.
For efficient OLED operation, the HOMO and LUMO
energy barriers between blocking and emissive layer need to be
small to guarantee efficient charge injection and a balanced
charge ratio. The HOMO and LUMO energies in 4P-NPD
films are −5.7 and −2.3 eV11, respectively, which is well
compatible with the blocker materials used in this study
(LUMOBPhen = −2.9 eV and HOMOSpiro‑TAD = −5.4 eV). As
discussed in more detail in Section 4.2, cyclic voltammetry
(CV) and DFT calculations both indicate that the increased
dihedral angle leads to minor changes in the HOMO position
(Table 2). Consequently, the exchange of the blue emitter
should only have marginal influence on hole injection into the
emitter layer.
In contrast, according to our DFT calculations, the vertical
electron affinity (EA) is expected to decrease along our series
(4P-NPD: 0.38 eV, 8M-4P-NPD: 0.24 eV, 8M-4P-FPD: 0.03
eV), indicating that electron injection from the HBL side into
either 8M-4P-NPD or 8M-4P-FPD may be impeded. However,
the computed values for IP and EA neglect effects arising from
embedding the molecule into a polarizable medium which is
expected to lead to substantial shifts. Changes of the vacuum
level in molecular thin films as a result of specific morphology
are also neglected. Compare ref 28 for details of the
polarization shift, ref 29 for typical values of the polarization
Figure 4. Synthesis route of 8M-4P-NPD and 8M-4P-FPD.
The quarterphenylene synthesis started from 4,4′-diiodo-
2,2′,6,6′-tetramethyl-1,1′-biphenyl (1).31 By using Buchwald−
Hartwig amination conditions the diiodo compound 1 was
converted to the monosubstituted product 2a in 71% yield, and
2b in 80% yield, respectively. Using an excess of 1 suppresses
the disubstitution and thus increases the yield of the desired
product, compared to applying equimolar quantities. The final
2417
dx.doi.org/10.1021/cm500602y | Chem. Mater. 2014, 26, 2414−2426

Chemistry of Materials
Article
Figure 5. Crystalline structure of 8M-4P-FPD. The rotation angles φ between the phenyl groups is φ (P1/P2) = φ (P3/P4) = 85.37°.
quarterphenylene compounds 8M-4P-NPD and 8M-4P-FPD
were synthesized in dimethylformamid (DMF) by copper-
mediated aryl−aryl homo coupling at elevated temperatures,
achieving good yields. Various temperature conditions and
reaction times were applied in coupling tests with 2a, and
stirring at 250 °C for 40 h was found to be the most efficient
procedure for selective quarterphenylene formation. Side
reaction products from reductive dehalogenation were also
observed, but these reaction products were not isolated due to
their similar retention on silica during chromatography. Both
new compounds showed good sublimation yields (∼ 70%) and
could be deposited conveniently under ultrahigh vacuum
atmosphere at evaporation temperatures between 240 and
280 °C.
quarterphenylene compounds. Within the stability window of
the solvent, one reversible oxidation process was observed for
both of the newly synthesized compounds and for the 4P-NPD
reference (Figure 6). The CV data and the resulting ionization
4. CHARACTERIZATION
4.1. Crystal Structure. The molecular geometry crystalline
structure of 8M-4P-FPD is derived from X-ray diffraction
(XRD) measurements on 8M-4P-FPD crystals grown from
toluene solution (Figure 5). The crystals belong to the
noncentrosymmetric space group Fdd2 with Z = 8 and contain
two toluene molecules per 8M-4P-FPD unit. The orthorhombic
unit cell has lattice vectors of a = 56.136(1) Å, b = 12.316(3) Å,
and c = 19.762(4) Å.
Figure 6. Cyclic voltammetry of 4P-NPD, 8M-4P-NPD, and 8M-4P-
FPD with respect to a Ag/AgCl electrode.
The two halves of the 8M-4P-FPD molecule occupy
symmetry equivalent positions generated by a 2-fold rotation
axis. The nitrogen atoms of the triarylamine system exhibit an
almost trigonal planar environment indicated by the angular
sum of 359.5°. The quarterphenylene backbone in the
molecular structure of 8M-4P-FPD is not linear but slightly
bent. Due to steric repulsion of the methyl groups, the relevant
phenylene rings P1 and P4 are arranged orthogonally to P2 and
P3 with an angle of 85.4°, as intended. The central biphenylene
unit exhibits a rotation angle of 41° between P2 and P3. The
dihedral angles realized in the crystal structure remain close to
the values computed for the free molecule, compare Table 1.
This good agreement indicates that the steric hindrance
introduced by the methyl groups has a larger impact on the
total energy than embedding the molecule into a crystal with
rather small intermolecular interactions (relying on weak van
der Waals forces).
potentials are summarized in Table 2. The values are in
reasonable agreement with DFT values of the ionization
potential, defined as the difference between the energies of a
cationic and a neutral molecule, and the respective polarization
shifts deduced from the ionic radii.
An average offset of 0.55 eV between ionization potentials
based on CV measurements and on DFT calculations plus
polarization shift may result from a systematic offset of our
ferrocene reference, or from the assumption of embedded
spherical ions when estimating the polarization shift. In
summary, based on CV and on DFT, we conclude that the
ionization potentials of the three model compound deviate by
less than 0.2 eV from each other, so that hole injection into the
emissive layer of OLEDs comprising either of the three
materials should be similar.
4.3. Spectroscopic Properties. Photoluminescence.
Photoluminescence (PL) measurements of films consisting of
our model compounds are shown in Figure 7. The emission
peaks of 4P-NPD, 8M-4P-NPD, and 8M-4P-FPD are at
λmax = 428, 418, and 390 nm, respectively. In the new
compounds, the prominent vibronic structure observed for
4P-NPD is blurred by a stronger broadening of each subband.
It was not possible to perform XRD measurements on 8M-
4P-NPD and 4P-NPD because these compounds do not form
single crystals.
4.2. Ionization Potential. Cyclic voltammetry in dichloro-
methane was used to investigate the electronic properties of the
2418
dx.doi.org/10.1021/cm500602y | Chem. Mater. 2014, 26, 2414−2426

Chemistry of Materials
Article
Figure 7. PL of 4P-NPD, 8M-4P-NPD, and 8M-4P-FPD thin film and
in toluene solution.
Figure 8. Phosphorescence spectra at low temperature (<100 K),
obtained on thin films of 4P-NPD, 8M-4P-NPD, and 8M-4P-FPD,
respectively. The PL spectrum of Ir(ppy)2(acac) at ambient temper-
ature is shown for comparison (gray curve).
The spectral shift toward smaller wavelengths leads to a
deeper blue emission color of the two new compounds
compared to 4P-NPD which exhibits Commission Internatio-
nale de l’Eclairage 1931 (CIE) color coordinates of (0.158/
0.074). The CIE coordinates of the new compounds show a
significantly reduced y-coordinate, (0.157/0.023) for 8M-4P-
NPD and (0.158/0.021) for 8M-4P-FPD. This, in turn, may
lead to a higher color rendering index in white OLEDs, which is
important for high-quality indoor lighting. However, the 8M-
4P-FPD emission spectrum partly extends to below 380 nm,
making this emitter less suitable for practical applications since
this part of the emission falls outside the visible range.
Both new compounds show only small spectral deviations
between their emission in thin film and in toluene solution,
indicating no significant molecular interaction between the
molecules even when densely packed. In contrast, the thin film
spectra of 4P-NPD are red-shifted by about 10 nm with respect
to the solution spectra. This difference in peak emission and the
broadening of the solution spectrum are attributed to a larger
influence of the surroundings onto 4P-NPD with the more
flexible central quarterphenyl chain, allowing for strong
elongations of low-frequency torsional modes.
Furthermore, we have measured the photoluminescence
quantum yield (PL-QY) of thin films of 4P-NPD, 8M-4P-NPD,
and 8M-4P-FPD to be 41 13%, 42 13%, and 14 4%,
respectively. Apparently, 8M-4P-NPD can be as efficient as 4P-
NPD when used as a bulk emitter within an OLED structure.
The low PL-QY of 8M-4P-FPD is not understood so far
because the additional fluorenyl groups were expected to
increase the fluorescence yield.
Phosphorescence. To achieve TH in OLEDs, the position
of the triplet level of the blue emitter is crucial. The triplet
energies of all studied compounds are derived from gated PL
measurements of thin films at low temperature (Figure 8). For
8M-4P-NPD and 8M-4P-FPD, the entire fluorescence spectra
are shifted toward smaller wavelengths, indicating that also the
triplet state should occur at a higher energy than in 4P-NPD. In
comparison with 4P-NPD, the phosphorescence spectra of the
two new compounds are broadened (cf. Figure 7). The
emission peaks lie at 566, 546, and 519 nm for 4P-NPD, 8M-
4P-NPD, and 8M-4P-FPD, respectively.
difference between the phosphorescence peaks obtained using
these two methods: interactions with the polystyrene, or
different relaxation dynamics in the triplet density of states.
Note that triplet energies reported elsewhere show similar
deviations. For example, for the well-known material N,N′-
di(naphthalene-1-yl)-N,N′-diphenyl-benzidine (NPD), which
chemically differs from 4P-NPD by just two phenyl rings,
values within the range between 514 and 563 nm have been
reported.13,32,33
For 8M-4P-NPD and 8M-4P-FPD, both the fluorescence and
the phosphorescence spectra are shifted toward shorter
wavelengths with respect to 4P-NPD, indicating that TH by
an efficient green phosphorescent emitter like Ir(ppy)2(acac)
should be within reach.
Analysis of Observed Line Shapes. For a meaningful
comparison between the average emission energies of different
molecules, the observed emission spectra I(λ) have to be
reported as a function of energy in the form I(E) ∝ I(λ)/E2.
Moreover, the proportionality of the emission intensity I(E) to
the density of photons scales as E3, requiring a further division
by E3 before the underlying molecular properties emerge.34
Hence, the vertical transition energy from the relaxed excited
geometry of a molecule can be determined by averaging over
the observed emission intensity, rescaled as I(E)/E3 ∝ I(λ)/E5.
This allows for a direct comparison with calculated transition
energies.
In practice, the rescaled emission intensity I(E)/E3 is fitted to
a Poisson progression over an effective internal mode ℏω with
transition probabilities:
P(S) = e−SSn/n!
(1)
n
toward the n-th vibrational level in the ground state potential,
including a different Gaussian broadening for each subband.
This allows defining the vertical transition energy Evert
E00 − Sℏω of either phosphorescence or fluorescence, where
00 is the apparent origin of the emission spectrum, and S is its
=
E
Huang−Rhys factor serving as the argument of the Poisson
progression. The data are summarized in Table 3. This
procedure provides meaningful values for the vertical transition
energies T1→S0 and S1→S0, even in cases where the long
wavelength part of the emission spectra extends beyond the
detection window.
The phosphorescence peak wavelength 566 nm of 4P-NPD
obtained within this work is considerably higher than the value
of 539 nm measured by Schwartz et al.11 Schwartz et al. doped
4P-NPD into a polystyrene matrix, while in this work a neat
film is used. Two reasons may account for the substantial
The vertical transition energies defined in this way reveal that
the observed splitting ΔST between the PL transition
2419
dx.doi.org/10.1021/cm500602y | Chem. Mater. 2014, 26, 2414−2426

Chemistry of Materials
Article
Table 3. Phosphorescence Line Shapes Analyzed in Terms of
a Poisson Progression over an Effective Internal Vibration
According to Eq 1; See Text for Further Details
5. BLUE OLEDS
To investigate the new materials and, further, to test their
potential as blue fluorescent bulk emitters, we fabricated
OLEDs with the structures shown in Figure 9a. The transport
layer thickness is identical for all three OLEDs to ensure optical
comparability. Different blocker materials are applied to the
different emitters since these were found to provide superior
performance than 4,7-diphenyl-1,10-phenanthroline (BPhen)
or 2,2′,7,7′-tetrakis-(N,N-diphenylamino)-9,9′-spirobifluorene
(Spiro-TAD) which are typically used in combination with
4P-NPD.12,20,22
4P-NPD
8M-4P-NPD
8M-4P-FPD
E00 (eV)
2.17
0.170
1.56
1.91
1.98
2.26
0.168
1.71
1.98
2.05
2.41
0.161
1.60
2.15
2.22
ℏωeff (eV)
Seff
T1→S0 = Evert (eV)
Eph (eV)
Compared to the 4P-NPD based OLED, the new
compounds show a much flatter current−voltage−luminance
(IVL) characteristic (Figure 9b), which may result from energy
barriers between the different layers and a charge imbalance.
This is in good agreement with the fact that luminance sets in
at higher voltages for 8M-4P-NPD and 8M-4P-FPD than for
4P-NPD. Figure 9c shows the normalized spectral emission of
our OLEDs (at a current density of 15.4 mA cm−2). These
spectra correlate with the PL measurements (cf. Figure 7)
confirming that the emission results from the respective emitter
material and not from any of the blocking layers. However, the
EQE of the OLEDs comprising the new compounds is strongly
reduced compared to the 4P-NPD OLED (Figure 9d). The
maximum EQEs for 8M-4P-NPD and 8M-4P-FPD OLEDs are
1.3% and 0.5%, respectively. This efficiency drop mainly results
from the higher LUMO energy of the new materials, which
hinders electron injection and leads to a deterioration of the
charge balance. In the case of 8M-4P-FPD, the lower PL-QY is
an additional reason for the reduced EQE.
S1→S0(S1) and the phosphorescence transition T1→S0(T1)
increases from 0.77 eV for 4P-NPD to 0.83 eV for 8M-4P-NPD
and 0.85 eV for 8M-4P-FDP (Table 4). Due to the definition of
each transition energy relying on an average over the rescaled
emission spectrum, for 4P-NPD the singlet−triplet splitting
ΔST = 0.77 eV does not match the previous estimate of 0.61 eV
deduced from the peak wavelengths.11
The calculated and observed values for the Stokes shift
between absorption, S1←S0 (S0), and phosphorescence,
T1→S0 (T1), compare favorably. The last two rows of
Table 4 give the difference between either calculated and
measured absorption energies, or calculated and measured
phosphorescence bands. For the three model compounds,
TD-DFT on average underestimates the observed vertical
transition energies by 0.16 eV.
For 4P-NPD, the S1 singlet excitation in the S0 geometry
resulting from the TD-DFT calculation amounts to 3.13 eV.
The T1 triplet emission from the optimized T1 geometry has an
energy of 1.71 eV, resulting in a calculated Stokes shift of 1.42
eV between the lowest absorption band and the average
phosphorescence. This value is in excellent agreement with the
measured energy difference of 1.43 eV. When comparing the
lowest measured absorption band and the phosphorescence
spectra for all three compounds, the Stokes shift is on average
1.42 eV, again in very good agreement with the calculated value
of 1.44 eV. This indicates that the energy difference between
the lowest absorption band and the phosphorescence energy
can be reproduced quite precisely with TD-DFT.
6. TH IN OLEDS
6.1. 8M-4P-NPD. As previously discussed, it is expected that
the triplet energy of 8M-4P-NPD is too low to allow efficient
triplet harvesting by green phosphors. Nevertheless, a test series
of TH OLEDs is fabricated. Figure 10a shows a proposed
energy scheme of the EML and the surrounding blocking
layers. It is assumed that the exciton generation zone, as in the
case of 4P-NPD, is close to the HBL interface, here TPBI, so
Table 4. Summary of Observed and Calculated Molecular Transition Energies, in eV; Starting Electronic Configuration and
Relaxed Geometry before the Transition Occurs Is Given in Parentheses; All Calculated TD-DFT Transition Energies Rely
Either on the Relaxed S0 Geometry or the Relaxed T1 Geometry
4P- NPD
8M-4P-NPD
8M-4P-FPD
Ir(ppy)2(acac)
calc
S1←S0 (S0)
T1→S0 (T1)
exp
3.13
1.71
3.21
1.74
3.49
2.06
2.77
2.18
S1←S0 (S0)
S1→S0 (S1)
T1→S0 (T1)
calc
3.34
2.68
1.91
3.44
2.81
1.98
3.51
3.00
2.15
2.21
S1←S0 (S0) − T1→S0 (T1)
exp
1.42
1.43
1.47
1.46
1.43
1.36
0.59
S1←S0 (S0) − T1→S0 (T1)
exp ΔST
S1→S0 (S1) − T1→S0 (T1)
calc − exp
0.77
0.83
0.85
S1←S0 (S0)
−0.21
−0.19
−0.23
−0.24
−0.02
−0.09
calc − exp
T1→S0 (T1)
−0.03
2420
dx.doi.org/10.1021/cm500602y | Chem. Mater. 2014, 26, 2414−2426

Chemistry of Materials
Article
Figure 9. (a) Layer architecture, (b) IVL-characteristics, (c) spectral radiance, and (d) EQE of the OLEDs comprising 4P-NPD, 8M-4P-NPD, and
8M-4P-FPD as EMLs.
that triplets can diffuse toward the phosphorescent emitter,
which is doped into 8M-4P-NPD at the other side of the EML.
Additionally, we vary the intrinsic layer thickness of 8M-4P-
NPD between 3 and 9 nm, which corresponds to the distance x
between exciton generation zone and TH zone. In this
experiment, the phosphorescent emitter Ir(ppy)2(acac) is
doped into the emitter to harvest triplets.
To avoid triplet quenching at the electron blocking layer, the
triplet level of Spiro-TAD needs to be higher than that of
Ir(ppy)2(acac). Because Spiro-TAD is typically used as blocker
material for green OLEDs, we assume that the triplet energy is
sufficiently high. To the best of our knowledge, the triplet
energy of Spiro-TAD was never determined experimentally.
The current−voltage (IV) characteristics are slightly dependent
on the 8M-4P-NPD intrinsic layer thickness x (Figure 10b); the
thicker the layer, the flatter the IV curve. This results from the
higher voltage drop over the intrinsic layer.
More importantly, the luminance is strongly increased by a
decrease of x. This is accompanied by a change in the emission
spectra (Figure 10c), where the green emission, which
dominates the luminance, increases with decreasing x. This
emission increase indicates typical TH behavior: Due to the
diffusion of triplets away from the exciton generation interface
and the corresponding exponential decrease of the number of
triplets with larger x, a small distance between exciton
generation zone and TH zone is most beneficial for triplet
harvesting. However, at the same time, the decrease of layer
thickness x leads to a loss in blue 8M-4P-NPD emission due to
singlet transfer from 8M-4P-NPD to Ir(ppy)2(acac). Here,
singlets are converted into triplets due to the high intersystem
crossing rate of Ir(ppy)2(acac) and additionally contribute to
the green emission.
The decrease in blue emission and increase in green emission
with increasing layer thickness x confirms the assumption that
the exciton generation zone is next to the HBL. With increasing
applied voltage the blue emission peak grows considerably
faster than the green emission, leading to a pronounced
voltage-dependence of the electroluminescence spectra. This
behavior is typical for white TH OLEDs, as the exciton density
on phosphorescent emitters is higher than on the fluorescent
emitter; their emission thus is more prone to bimolecular
quenching.
At low luminance, the EQE of the TH OLEDs (Figure 10d)
is improved by more than a factor of 2 compared to the
previously discussed blue 8M-4P-NPD OLEDs (cf. Figure 9d).
These results suggest that Ir(ppy)2(acac) emission is caused by
TH, while contributions resulting from direct recombination
and singlet transfer (at small distances) cannot be fully
excluded. This is interesting considering that according to
Table 4, the triplet transition of 8M-4P-NPD lies 0.23 eV below
that of Ir(ppy)2(acac). One possible explanation for this
behavior is that the 8M-4P-NPD phosphorescence sets in at
509 nm (cf. Figure 8), implying that the fraction of triplets at
the high energy end of the density of states transfer to the
triplet level of Ir(ppy)2(acac), where they decay radiatively.
However, the blue reference OLEDs cannot provide an EQE
above 1.3%, which limits the overall EQE of the 8M-4P-NPD
TH OLEDs. For comparison, OLEDs with only 4P-NPD as
emitter achieve EQE values in the range of 4%.23 Because 4P-
NPD and 8M-4P-NPD have a similar PL-QY, we assume that
2421
dx.doi.org/10.1021/cm500602y | Chem. Mater. 2014, 26, 2414−2426

Chemistry of Materials
Article
Figure 10. (a) Energy scheme, (b) IVL-characteristics, (c) spectral radiance, and (d) EQE of the 8M-4P-NPD TH OLEDs.
the difference in EQE is caused by the hindered electron
injection into 8M-4P-NPD: The LUMO energy of 8M-4P-
NPD of −1.8 eV provides a high energy barrier for electrons to
the HBL TPBI of −1.0 eV.
6.2. 8M-4P-FPD. A similar series of devices was studied for
the other blue emitter 8M-4P-FPD. Figure 11 illustrates the
performance of the TH OLEDs based on 8M-4P-FPD, with a
distance variation of x = 3, 5, and 7 nm. In comparison to 8M-
4P-NPD, we find the same trends in IVL behavior (Figure
11b), spectral emission (Figure 11c), and EQE (Figure 11d)
upon increasing x. However, the contribution of Ir(ppy)2(acac)
emission to the spectrum is much higher compared to the case
of 8M-4P-NPD.
This is in good agreement with the fact that due to the higher
triplet energy of 8M-4P-FPD, a larger fraction of triplets can be
transferred to Ir(ppy)2(acac). Singlet exciton transfer is only
observed for small distances x as in the case of 8M-4P-NPD,
reducing in turn the blue singlet emission, e.g. for x = 3 nm.
For the TH OLED with 3 nm intrinsic 8M-4P-FPD
thickness, the EQE approaches 9%. Despite the apparent high
efficiency of TH from 8M-4P-FPD, the blue emission remains
rather weak. Again, the high LUMO energy of −1.7 eV hinders
the efficient electron injection into 8M-4P-FPD, thus
preventing a balanced distribution of holes and electrons.
small changes of the structure of 4P-NPD: Quantum chemical
calculations indicate that an increase in triplet energy occurs by
increasing the rotation angle of the outer phenyl rings (8M-4P-
NPD), and, further, by the incorporation of fluorenyl side
groups (8M-4P-FPD). This prediction was verified exper-
imentally by measuring and analyzing the phosphorescence
spectra of 4P-NPD, 8M-4P-NPD, and 8M-4P-FPD. Vertical
transition energies for the T1→S0 transition as high as 2.15 eV
were observed for 8M-4P-FPD.
The two new compounds 8M-4P-NPD and 8M-4P-FPD are
implemented as blue bulk emitters into OLEDs. A maximum
EQE of 1.3% and 0.5% is observed for 8M-4P-NPD and 8M-
4P-FPD, respectively. This is considerably lower than the 3.4%
reached by 4P-NPD due to the high LUMO levels of the new
compounds, which impede electron injection and lead to
charge imbalance as well as high driving voltages. To overcome
the issue of electron injection, additional electron withdrawing
groups at the molecule (such as bromide) might be helpful to
shift the HOMO and LUMO level toward lower energies, while
maintaining the triplet energy. Thereby, charge carrier balance
and OLED efficiencies can be improved.15
Finally, we demonstrated triplet harvesting by the green
phosphorescent emitter Ir(ppy)2(acac) for both new com-
pounds. A maximum of 9% EQE was obtained for the blue-
green TH OLED comprising 8M-4P-FPD, which is basically
limited by the efficiency of the blue emitter. However, our
energy considerations using peak wavelength or average energy
of the phosphorescence spectrum cannot entirely explain TH
by the green emitter. Nevertheless, this research provides a
7. SUMMARY AND CONCLUSIONS
In summary, we developed two blue fluorescent bulk emitters
with high triplet energies to facilitate triplet harvesting by green
phosphorescent emitters. Chemical engineering was based on
2422
dx.doi.org/10.1021/cm500602y | Chem. Mater. 2014, 26, 2414−2426

Chemistry of Materials
Article
Figure 11. (a) Energy scheme, (b) IVL-characteristics, (c) spectral radiance, and (d) EQE of the 8M-4P-FPD TH OLEDs.
precipitate was filtered off with a Buchner funnel and the solvent was
̈
useful strategy for engineering new blue bulk emitters that are
suitable for triplet harvesting with green phosphors.
removed with a rotary evaporator and dried in vacuum. Column
chromatography with cyclohexane/toluene (1:1) and silica (K60) was
applied for purification. The product (Rf = 0.52) was isolated as a
colorless solid (10.94 g, 38.33 mmol, 90%). mp 107 °C. C21H19N
(285.38 g mol−1): calcd (%) C 88.38, H 6.71, N 4.91; found C 88.59,
H 6.71, N 4.91. 1H NMR (CDCl3, 500 MHz) δ (ppm): 7.61 (br m, 2
H, Har), 7.39 (d, 3JHH = 7.6 Hz, 1 H, Har), 7.21−7.32 (br m, 4 H, Har),
7.11 (br m, 4 H, Har), 6.94 (br m, 1 H, Har), 5.80 (s, 1 H, NH), 1.46
(s, 6 H, CH3). 13C NMR (acetone-d6, 126 MHz) δ (ppm): 155.9 (C),
153.8 (C), 144.7 (C), 144.2 (C), 140.3 (C), 132.5 (C), 130.0 (CH),
127.7 (CH), 128.3 (CH), 127.1 (CH), 126.7 (CH), 123.2 (CH),
121.6 (CH), 120.9 (CH), 119.7 (CH), 117.8 (CH), 117.1 (CH),
112.4 (CH), 47.3 (C), 27.5 (CH3). ESI-MS (+10 V, MeOH, 0.1%
NH4OAc): m/z 286.3 [M + H]+.
N-(4′-Iodo-2,2′,6,6′-tetramethyl-[1,1′-biphenyl]-4-yl)-N-phenyl-
naphthalen-1-amine (2a). In a flame-dried Schlenk flask 4,4′-diiodo-
2,2′,6,6′-tetramethyl-1,1′-biphenyl (8.35 g, 18.07 mmol, 200 mol %)
and palladium(II)acetate (61 mg, 0.27 mmol, 3 mol %) were dissolved
in dry toluene (180 mL). After addition of tri-tert-butylphosphine
solution in toluene (1.4 mL, 1 mol L−1, 1.4 mmol, 15 mol %) the color
turned from yellow to orange and the solution was continuously
stirred. Within 15 min, the color disappeared and sodium tert-butoxide
(1.04 g, 10.82 mmol, 120 mol %) were added in one portion. The
mixture was heated in an oil bath (105 °C) and N-naphthylaniline
(1.98 g, 9.03 mmol) dissolved in toluene (30 mL) was added
dropwise. During the addition, a white solid of sodium iodide
precipitated. Stirring and heating continued overnight. The precipitate
8. EXPERIMENTAL DETAILS
General Remarks. Buchwald−Hartwig amination reactions were
carried out using the standard Schlenk technique under nitrogen.
Commercially available reagents and solvents were used as received.
Extra dry toluene over molecular sieve was purchased from Acros. Tri-
tert-butylphosphine (abcr GmbH, Germany) was diluted with dry
toluene to obtain a solution with a concentration of 1 mol L−1. The
C−C coupling in N,N-dimethylformamide (for analysis, AppliChem
GmbH) was performed in an Ace Glass pressure tube with copper
bronze, which was activated with sodium EDTA and stored under
nitrogen atmosphere. 1H and 13C NMR spectra were recorded at
25 °C with a Bruker DRX 500 P instrument at 500.13 and 125.76
MHz, respectively. The assignment of quaternary C, CH, CH2, and
CH3 was completed using DEPT spectra. Elemental analyses were
achieved with a Eurovektor Hekatech EA-3000 elemental analyzer. All
mass spectra were recorded with a Bruker Esquire-LC 00084
instrument. The quarterphenylene compounds were purified in
vacuum using a gradient sublimation device purchased from CreaPhys
GmbH, Dresden. The sublimation yield given for the quarter-
phenylene compounds is the overall yield of two sublimations.
9,9-Dimethyl-N-phenyl-9H-fluoren-2-amine. In a flame-dried
Schlenk flask 2-bromo-9,9-dimethyl-9H-fluoren (11.63 g, 42.57 mmol)
and palladium(II)acetate (287 mg, 1.28 mmol, 3 mol %) were
dissolved in dry toluene (250 mL).35 After addition of tri-tert-
butylphosphine solution in toluene (6.4 mL, 1 mol L−1, 6.4 mmol,
15 mol %) the color turned from yellow to orange and the solution
was further stirred. During 15 min the color disappeared and sodium
tert-butoxide (6.1 g, 63.48 mmol, 150 mol %) was added in one
portion, followed by aniline (7.8 mL, 85.43 mmol). The mixture was
heated in an oil bath (105 °C) while stirring overnight. The white
was filtered off with a Buchner funnel and the solvent was removed
̈
with a rotary evaporator and dried in vacuum. Column chromatog-
raphy with hexane/chloroform (4:1) and silica (K60) was applied to
recover unreacted diiodo-tetramethyl-biphenyl and isolate the product
(3.50 g, 6.56 mmol, 71%, Rf = 0.44). mp 140 °C. C32H28IN (553.48 g
2423
dx.doi.org/10.1021/cm500602y | Chem. Mater. 2014, 26, 2414−2426

Chemistry of Materials
Article
mol−1): calcd (%) C 69.44, H 5.10, N 2.53; found C 70.47, H 5.10, N
2.83. 1H NMR (CDCl3, 500 MHz) δ (ppm): 7.97 (d, 3JHH = 8.5 Hz, 1
H, Har), 7.88 (d, 3JHH = 8.2 Hz, 1 H, Har), 7.76 (d, 3JHH = 8.2 Hz, 1 H,
Har), 7.44−7.50 (m, 2 H, Har), 7.46 (s, 2 H, CH), 7.33−7.37 (m, 2 H,
Har), 7.16−7.20 (m, 2 H, Har), 6.98−7.00 (m, 2 H, Har), 6.88−6.92
(m, 1 H, Har), 6.80 (s, 2 H, CH), 1.88 (s, 6 H, CH3), 1.72 (s, 6 H,
CH3). 13C NMR (CDCl3, 126 MHz) δ (ppm): 148.7 (C), 147.1 (C),
143.7 (C), 139.6 (C), 138.6 (C), 136.2 (CH), 136.1 (C), 135.3 (C),
132.7 (C), 131.3 (C), 129.0 (CH), 128.3 (CH), 127.1 (CH), 126.4
(CH), 126.2 (CH), 126.2 (CH), 126.0 (CH), 124.4 (CH), 121.4
(CH), 121.1 (CH), 92.4 (CI), 20.0 (CH3), 19.4 (CH3). ESI-MS (+20
V, MeOH, 0.1% NH4OAc): m/z 554.3.2 [M + H]+.
N-(4′-Iodo-2,2′,6,6′-tetramethyl-[1,1′-biphenyl]-4-yl)-9,9-dimeth-
yl-N-phenyl-9H-fluoren-3-amine (2b). In a flame-dried Schlenk flask
4,4′-diiodo-2,2′,6,6′-tetramethyl-1,1′-biphenyl (11.46 g, 24.80 mmol,
200 mol %) and palladium(II)acetate (84 mg, 0.37 mmol, 3 mol %)
were dissolved in dry toluene (300 mL). After addition of tri-tert-
butylphosphine solution in toluene (1.9 mL, 1 mol L−1, 1.9 mmol, 15
mol %) the color turned from yellow to orange and the solution was
further stirred. Within 15 min, the color disappeared and sodium tert-
butoxide (1.79 g, 18.63 mmol, 120 mol %) was added in one portion.
The mixture was heated in an oil bath (105 °C) and 9,9-dimethyl-N-
phenyl-9H-fluoren-3-amine (3.54 g, 12.40 mmol) in toluene (35 mL)
was added dropwise. During the addition, a white solid of sodium
iodide precipitated. Stirring and heating continued overnight. The
40 h. After cooling to ambient temperature, methanol (40 mL)
followed by water (40 mL) were added, and the precipitating white
solid was filtered off. Column chromatography with silica and
cyclohexane/toluene as eluent (2.5:1) was applied to obtain the
product as a colorless solid (0.92 g, 0.934 mmol, 67%, Rf = 0.46). mp
316 °C. C74H68N2 (985.35g mol−1): calcd (%) C 90.20, H 6.96, N
1
2.84; found C 89.98, H 7.01, N 3.02. H NMR (CDCl3, 500 MHz) δ
(ppm): 7.63 (d, 3JHH = 7.4 Hz, 1 H, Har), 7.59 (d, 3JHH = 8.2 Hz, 1 H,
3
Har), 7.45 (s, 2 H, CH), 7.39 (d, JHH = 7.4 Hz, 1 H, Har), 7.22−7.32
(m, 5 H, Har), 7.17−7.19 (m, 2 H, Har), 7.07−7.10 (m, 1 H, Har),
7.00−7.03 (m, 1 H, Har), 6.93 (s, 2 H, CH), 2.06 (s, 6 H, CH3), 1.85
(s, 6 H, CH3), 1.42 (s, 6 H, CH3). 13C NMR (CDCl3, 126 MHz) δ
(ppm): 154.9 (C), 153.5 (C), 148.1 (C), 147.5 (C), 146.2 (C), 139.3
(C), 139.1 (C), 138.7 (CH), 136.7 (C), 136.2 (C), 134.8 (C), 133.7
(C), 129.1 (CH), 126.9 (CH), 126.3 (CH), 126.0 (CH), 123.7 (CH),
123.4 (CH), 123.1 (CH), 122.4 (CH), 122.1 (CH), 120.5 (CH),119.3
(CH), 118.5/CH), 46.8 [C(CH3)], 27.1 (CH3), 20.1 (CH3), 20.0
(CH3). ESI-MS (+20 V, MeOH, 0.1% NH4OAc): m/z 985.8 [M +
H]+.
Single-Crystal X-ray Structure Determination of 8M-4P-FPD.
Single crystals suitable for the X-ray diffraction study were obtained
from toluene solution. The crystallographic data were collected on a
Bruker Nonius Kappa CCD diffractometer using graphite-monochro-
mated Mo−Kα (λ = 0.71073 Å) radiation at 193 K. SADABS36 was
used to perform area-detector scaling and absorption corrections. The
structures were solved by direct methods using Sir9737 and refined by
full-matrix least-squares techniques against Fo2 by using SHELXL-97.38
All non-hydrogen atoms were refined anisotropically. Hydrogen atom
positions were generated by their idealized geometry and refined
within a riding model. The absolute structure was determined from
anomalous dispersion effects and verified according to Flack.39
Cyclic voltammetry was recorded on a Metrohm μ-Autolab
instrument in a single-component cell under a nitrogen atmosphere.
A typical three electrode configuration with an inlaid platinum disk as
working electrode, platinum wire as counter electrode, and a silver rod
coated electrochemically with AgCl was used. The measurements were
performed with a scan rate of 100 mV s−1 in degassed dry
dichloromethane (HPLC quality) with tetra-n-butylammonium
hexafluorophosphate (TBAPF, recrystallized from ethanol,
0.1 mol L−1) as electrolyte.
As an absolute reference, the oxidation potential of ferrocene was
determined in the same solvent, giving 0.40 V. Polarization shifts of all
compounds were estimated from the radius of each ion in the form28
P+ = e2/(4π0rion) (1 − 1/ε). For the three model compounds, the ionic
radii were estimated by assuming that their density coincides with the
crystalline phase of 8M-4P-FPD, and by equating the volume per
molecule with a sphere of radius rion. For ferrocene, a volume per
molecule of V0 = 205 Å3 was deduced from the respective crystal
phase,40 giving rion = 3.66 Å, and together with the dielectric constant ε
= 8.93 of dichloromethane, for Fc/Fc+ the polarization shift of the
ionization potential amounts to P+ = 1.75 eV.
Together with the ionization potential of a free ferrocene molecule
of IP = 6.64 eV,41 the modified ionization potential in dichloro-
methane occurs at 4.89 eV below vacuum, placing the reference
voltage 0 V of our electrochemical cell at 4.49 V below vacuum.
Fluorescence spectra were taken with a FSP920 fluorescence
spectrometer (Edinburgh Instruments). For Ir(ppy)2(acac), 8 wt % of
Ir(ppy)2(acac) were doped into a TCTA matrix to avoid concentration
quenching.
Phosphorescence. For phosphorescence (gated luminescence)
measurements, samples were excited with a pulsed YAG laser emitting
at 355 nm (EKSPLA) at 45° angle to the substrate plane. Emission
was focused onto a spectrograph and detected on a gated iCCD
camera (Stanford Computer Optics) with sub nanosecond resolution.
Mercury Argon light source was used to calibrate wavelength position,
and NIST traceable Tungsten Halogen white light source was used to
calibrate camera spectral response. For low temperature measure-
ments, samples were placed in the displex cryostat.
precipitate was filtered off with a Buchner funnel and the solvent was
̈
removed with a rotary evaporator and dried in vacuum. Column
chromatography with hexane/chloroform (2.5:1) and silica (K60) was
applied to recover unreacted diiodo-tetramethyl-biphenyl and isolate
the product (6.16 g, 9.94 mmol, 80%, Rf = 0.47). mp 185 °C.
C37H34IN (619.58 g mol−1): calcd (%) C 71.73, H 5.53, N 2.26; found
1
C 72.42, H 6.09, N 2.22. H NMR (CDCl3, 500 MHz): δ = 7.63 (d,
3JHH = 7.3 Hz, 1 H, Har), 7.58 (d, 3JHH = 8.0 Hz, 1 H, Har), 7.49 (s, 2
3
H, CH), 7.38 (d, JHH = 8.2 Hz, 1 H, Har), 7.23−7.32 (m, 4 H, Har),
7.14−7.19 (m, 3 H, Har), 6.99−7.07 (m, 2 H, Har), 6.89 (s, 2 H, CH),
1.92 (s, 6 H, CH3), 1.78 (s, 6 H, CH3), 1.41 (s, 6 H, CH3). 13C NMR
(CDCl3, 126 MHz) δ (ppm): 154.9 (C), 153.5 (C), 148.0 (C), 147.4
(C), 146.6 (C), 139.6 (C), 139.0 (C), 138.5 (C), 136.3 (C), 136.2
(CH), 133.9 (C), 133.5 (C), 129.1 (CH), 126.9 (CH), 126.4 (CH),
123.8 (CH), 123.2 (CH), 122.4 (CH), 122.3 (CH), 120.5 (CH),
119.3 (CH), 118.6 (CH), 92.4 (C), 46.7 (C), 27.0 (CH3), 19.9 (CH3),
19.5 (CH3). ESI-MS (+10 V, MeOH, 0.1% NH4OAc): m/z 620.2 [M
+ H]+.
2,2′,2‴,3″,5″,6,6′,6‴-Octamethyl-N4,N4‴-di(naphthalen-1-yl)-
N4,N4‴-diphenyl-[1,1′:4′,1″:4″,1‴-quaterphenyl]-4,4‴-diamine
(8M-4P-NPD). A pressure tube was charged with compound 2a (524
mg, 0.982 mmol), copper bronze (254 mg), and degassed DMF (1.0
mL). The tube was sealed and heated to 250 °C for 40 h under
stirring. After cooling to ambient temperature, 10 mL of methanol
followed by water (10 mL) were added, and the precipitating white
solid was filtered off. Column chromatography with cyclohexane/
toluene as eluent (2.5:1) was applied to obtain the product as a
colorless solid (283 mg 0.332 mmol, 68%, Rf = 0.47). mp 305 °C.
C64H56N2 (853.14g mol−1): calcd (%) C 90.10, H 6.62, N 3.28; found
C 90.16, H 6.67, N 3.15. 1H NMR (CDCl3, 500 MHz) δ (ppm): 8.01
(d, 3JHH = 8.4 Hz, 1 H, Har), 7.89 (d, 3JHH = 8.2 Hz, 1 H, Har), 7.77 (d,
3JHH = 8.2 Hz, 1 H, Har), 7.40 (s, 2 H, CH), 7.45−7.52 (m, 2 H, Har),
7.36−7.40 (m, 2 H, Har), 7.18−7.21 (m, 2 H, Har), 7.01−7.03 (m, 2 H,
Har), 6.89−6.92 (m, 1 H, Har), 6.87 (s, 2 H, CH), 2.01 (s, 6 H, CH3),
1.80 (s, 6 H, CH3). 13C NMR (CDCl3, 126 MHz) δ (ppm): 148.9
(C), 146.8 (C), 143.9 (C), 139.2 (C), 138.7 (C), 136.4 (CH), 136.3
(C), 135.3 (C), 131.4 (C), 128.9 (CH), 128.2 (CH), 127.1 (CH),
126.4 (CH), 126.1 (CH), 126.0 (CH), 124.5 (CH), 121.6 (CH),
121.2 (CH), 120.8 (C), 20.0 (CH3). ESI-MS (+10 V, MeOH, 0.1%
NH4OAc): m/z 853.8 [M + H]+.
N4,N4‴-Bis(9,9-dimethyl-9H-fluoren-2-yl)-2,2′,2‴,3″,5″,6,6′,6‴-
octamethyl-N4,N4‴-diphenyl-[1,1′:4′,1″:4″,1‴-quaterphenyl]-4,4‴-
diamine (8M-4P-FPD). A pressure tube was charged with compound
2b (1.72 g, 2.78 mmol), copper bronze (759 mg), and degassed DMF
(2.8 mL). The tube was sealed and heated under stirring to 250 °C for
Photoluminescence Quantum Yield. (PL-QYs) of films were
measured using an integrated sphere and Jobin Yvon Fluoromax-3
2424
dx.doi.org/10.1021/cm500602y | Chem. Mater. 2014, 26, 2414−2426

Chemistry of Materials
Article
fluorimeter.42 System accuracy was checked by measuring sublimed
100 nm tris(8-quinolionolato)aluminum(III) films whose PL-QYs
have already been reported.43 Comparable values were obtained.
Quantum Chemical Calculations. Theoretical studies of the
molecular geometries and their transition energies have been
performed with a time-dependent self-consistent charge density-
functional theory based tight-binding method, TD-SCC-DFTB,44−48
and with the hybrid functional B3LYP. Both approaches agree
qualitatively, and all theoretical considerations discussed here are
based on geometry optimizations at the B3LYP/6-31G(d) level and on
transition energies calculated with time-dependent density functional
theory (TD-DFT) using the same functional and variational basis set.
OLED Fabrication. Organic materials were commercially purchased
and purified by vacuum gradient sublimation. Before device
fabrication, prestructured ITO-coated glass substrates (Thin Film
Devices Inc.) were cleaned using ultrasonic treatment in N-methyl-2-
pyrrolidon, distilled water, and ethanol. Organic and metal layers are
deposited by thermal evaporation in a UHV chamber (Kurt J. Lesker
Co.) at a base pressure of 10−7 mbar without breaking the vacuum.
Evaporation rates and thicknesses of all layers were measured in situ
via quartz crystals. Doping was achieved by coevaporation. After
processing, OLEDs were immediately encapsulated in nitrogen
atmosphere using glass lids and a UV-curing epoxy resin. The active
area of all devices is 6.49 mm2.
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Annette Petrich for material sublimation, the IAPP
Lesker team for device fabrication, and Novaled AG (Dresden)
for providing organic materials. Simonas Krotkus is acknowl-
edged for fruitful discussion. The work leading to these results
has received funding from the European Community’s Seventh
Framework Programme under grant FP7-224122
(OLED100.eu).
ABBREVIATIONS
■
CIE, Commission Internationale de l’Eclairage 1931; CV, cyclic
voltammetry; DFT, density functional theory; DMF, dimethyl-
formamid; EA, electron affinity; EML, emission layer; EBL,
electron blocking layer; ETL, electron transport layer; EQE,
external quantum efficiency; HBL, hole blocking layer; HOMO,
highest occupied molecular orbital; HTL, hole transport layer;
IP, ionization potential; IVL, current−voltage−luminance;
LUMO, lowest unoccupied molecular orbital; OLED, organic
light-emitting diode; PL, photoluminescence; PL-QY, photo-
luminescence quantum yield; TH, triplet harvesting; UHV,
ultrahigh vacuum; XRD, X-ray diffraction
OLED Stacks. The layer structure of the 4P-NPD TH OLEDs (cf.
Figure 1b) was glass (1.1 mm)/ ITO (90 nm)/MeO-TPD:F6-
TCNNQ (55 nm, 2 wt %)/Spiro-TAD (10 nm)/4P-NPD:Ir-
(MDQ)2(acac) or 4P-NPD:Ir(ppy)2(acac) (5 nm, 5 wt %)/BPhen
(10 nm)/BPhen:Cs (35 nm)/Al (100 nm).
The layer structure of the three blue OLEDs (cf. Figure 9a) was (i)
glass (1.1 mm)/ITO (90 nm)/MeO-TPD:F6-TCNNQ (20 nm, 2 wt
%)/Spiro-TAD (10 nm)/4P-NPD (10 nm)/BPhen (10 nm)/
BPhen:Cs (35 nm)/Al (100 nm), (ii) glass (1.1 mm)/ITO (90
nm)/MeO-TPD:F6-TCNNQ (20 nm, 2 wt %)/Spiro-TAD (10 nm)/
8M-4P-NPD (10 nm)/TPBI (10 nm)/BPhen:Cs (35 nm)/Al (100
nm), and (iii) glass (1.1 mm)/ITO (90 nm)/MeO-TPD:F6-TCNNQ
(20 nm, 2 wt %)/TAPC (10 nm)/8M-4P-FPD (10 nm)/SPPO1 (10
nm)/BPhen:Cs (35 nm)/Al (100 nm). The green−blue TH OLEDs
consisted of (i) glass (1.1 mm)/ITO (90 nm)/MeO-TPD:F6-
TCNNQ (55 nm, 2 wt %)/Spiro-TAD (10 nm)/8M-4P-NPD:Ir-
(ppy)2(acac) (5 nm, 5 wt %)/8M-4P-NPD (3, 5, 7, or 9 nm)/TPBI
(10 nm)/BPhen:Cs (55 nm)/Al (100 nm) and (ii) glass (1.1 mm)/
ITO (90 nm)/ eO-TPD:F6-TCNNQ (55 nm, 2 wt %)/TAPC (10
nm)/8M-4P-FPD:Ir(ppy)2(acac) (5 nm, 5 wt %)/8M-4P-FPD (3, 5,
or 7 nm)/SPPO1 (10 nm)/BPhen:Cs (55 nm)/Al (100 nm).
OLED Characterization. All OLED characterizations were done in
air and under ambient temperature. Current−voltage−luminance
curves were measured using a source measure unit (Keithley SMU
2400) and a calibrated Si-photodiode. Spectral radiance was recorded
with a calibrated spectrometer (Instrument Systems GmbH CAS140).
Angular dependent measurements were performed with a self-made
and self-calibrated spectro-goniometer setup including an Ocean
Optics USB 4000 miniature fiber optic spectrometer. Efficiencies were
calculated using the absolute forward emission spectrum and the
angular dependent radiant intensity recorded by the goniometer.
MATERIALS ABBREVIATIONS
■
4p-NPD, N,N′-di-1-naphthalenyl-N,N′-diphenyl-
[1,1′:4′,1″:4″,1‴-quaterphenyl]-4,4‴ diamine; 8M-4P-NPD,
2,2′,2‴,3″,5″,6,6′,6‴-octamethyl-N4,N4‴-di(naphthalen-1-yl)-
N4,N4‴-diphenyl-[1,1′:4′,1″:4″,1‴-quaterphenyl]-4,4‴-dia-
mine; 8M-4P-FPD, N4,N4‴-bis(9,9-dimethyl-9H-fluoren-2-yl)-
2,2′,2‴,3″,5″,6,6′,6‴-octamethyl-N4,N4‴-diphenyl-
[1,1′:4′,1″:4″,1‴-quaterphenyl]-4,4‴-diamine; BPhen, 4,7-di-
phenyl-1,10-phenanthroline; F6-TCNNQ, 2,2′-(perfluoronaph-
thalene-2,6-diylidene)dimalono-nitrile; Ir(MDQ)2(acac),
iridium(III)bis(2-methyldibenzo-[f,h]chinoxalin)-
(acetylacetonate); Ir(ppy)2(acac), iridium(III)bis(2-phenylpyr-
idinato-N,C2′) acetylacetonate; ITO, indium tin oxide; MeO-
TPD, N,N,N′,N′-tetrakis(4-methoxyphenyl)-benzidine; Spiro-
TAD, 2,2′,7,7′-tetrakis-(N,N-diphenylamino)-9,9′-spirobifluor-
ene; SPPO1, N,N′-di(naphthalene-1-yl)-N,N′-diphenyl-benzi-
dine2-(diphenylphosphoryl)spirofluorene; TAPC, di-[4-(N,N-
ditolyl-amino)-phenyl] cyclohexane; TPBI, 2,2′,2″-(1,3,5-
phenylene)tris(1-phenyl-1H-benzimidazole)
REFERENCES
■
̈
AUTHOR INFORMATION
(1) Reineke, S.; Lindner, F.; Schwartz, G.; Seidler, N.; Walzer, K.;
Lussem, B.; Leo, K. Nature 2009, 459, 234.
(2) Hofmann, S.; Thomschke, M.; Lussem, B.; Leo, K. Opt. Express
2011, 19, 1250.
■
Corresponding Author
*E-mail: simone.hofmann@iapp.de.
̈
(3) Yersin, H.; Rausch, A. F.; Czerwieniec, R.; Hofbeck, T.; Fischer,
T. Coord. Chem. Rev. 2011, 255, 2622.
(4) Adachi, C.; Baldo, M. A.; Thompson, M. E.; Forrest, S. R. J. Appl.
Phys. 2001, 90, 5048.
(5) Sasabe, H.; Kido, J. Chem. Mater. 2011, 23, 621.
(6) Seifert, R.; Rabelo de Moraes, I.; Scholz, S.; Gather, M. C.;
Present Addresses
∥Novaled AG, Tatzberg 49, 01307 Dresden, Germany.
⊥Fraunhofer Institute for Photonic Microsystems IPMS, Maria-
Reiche-Str. 2, 01109 Dresden, Germany.
#Department of Physics, Kent State University, 105 Smith Hall,
Kent, Ohio 44242.
Lussem, B.; Leo, K. Org. Electron. 2013, 14, 115.
̈
$School of Physics and Astronomy, University of St Andrews,
North Haugh, St Andrews, KY16 9SS, Scotland, UK.
(7) Uoyama, H.; Goushi, K.; Shizu, K.; Nomura, H.; Adachi, C.
Nature 2012, 492, 234.
2425
dx.doi.org/10.1021/cm500602y | Chem. Mater. 2014, 26, 2414−2426

Chemistry of Materials
Article
(8) Jankus, V.; Chiang, C.-J.; Dias, F.; Monkman, A. P. Adv. Mater.
2013, 25, 1455.
(9) Sun, Y.; Giebink, N. C.; Kanno, H.; Ma, B.; Thompson, M. E.;
(43) Kawamura, Y.; Sasabe, H.; Adachi, C. Jpn. J. Appl. Phys. 2004,
43, 7729.
(44) Seifert, G.; Porezag, D.; Frauenheim, T. Int. J. Quantum Chem.
1996, 58, 185.
Forrest, S. R. Nature 2006, 440, 908.
(45) Porezag, D.; Frauenheim, T.; Kohler, T.; Seifert, G.; Kaschner,
R. Phys. Rev. B 1995, 51, 12947.
(10) Schwartz, G.; Pfeiffer, M.; Reineke, S.; Walzer, K.; Leo, K. Adv.
Mater. 2007, 19, 3672.
̈
(46) Elstner, M.; Porezag, D.; Jungnickel, G.; Elsner, J.; Haugk, M.;
Frauenheim, T.; Suhai, S.; Seifert, G. Phys. Rev. B 1998, 58, 7260.
(47) Niehaus, T.; Suhai, S.; Della Sala, F.; Lugli, P.; Elstner, M.;
Seifert, G.; Frauenheim, T. Phys. Rev. B 2001, 63, 085108.
(48) Niehaus, T. A. J. Mol. Struct.: THEOCHEM 2009, 914, 38.
(11) Schwartz, G.; Reineke, S.; Rosenow, T. C.; Walzer, K.; Leo, K.
Adv. Funct. Mater. 2009, 19, 1319.
(12) Rosenow, T. C.; Furno, M.; Reineke, S.; Olthof, S.; Lussem, B.;
̈
Leo, K. J. Appl. Phys. 2010, 108, 113113.
(13) Kondakova, M. E.; Deaton, J. C.; Pawlik, T. D.; Giesen, D. J.;
Kondakov, D. Y.; Young, R. H.; Royster, T. L.; Comfort, D. L.; Shore,
J. D. J. Appl. Phys. 2010, 107, 014515.
(14) Hung, W.-Y.; Chi, L.-C.; Chen, W.-J.; Chen, Y.-M.; Chou, S.-H.;
Wong, K.-T. J. Mater. Chem. 2010, 20, 10113.
(15) Brunner, K.; Van Dijken, A.; Borner, H.; Bastiaansen, J.; Kiggen,
̈
N.; Langeveld, B. J. Am. Chem. Soc. 2004, 126, 6035.
(16) Yang, X.; Zheng, S.; Bottger, R.; Chae, H. S.; Tanaka, T.; Li, S.;
Mochizuki, A. J. Phys. Chem. 2011, 115, 14347.
(17) Zheng, C.-J.; Wang, J.; Ye, J.; Lo, M.-F.; Liu, X.-K.; Fung, M.-K.;
Zhang, X.-H.; Lee, C.-S. Adv. Mater. 2013, 25, 2205−2211.
(18) Schwartz, G.; Reineke, S.; Walzer, K.; Leo, K. Appl. Phys. Lett.
2008, 92, 053311.
(19) Blochwitz, J.; Pfeiffer, M.; Fritz, T.; Leo, K. Appl. Phys. Lett.
1998, 73, 729.
(20) Hofmann, S.; Rosenow, T.; Gather, M. C.; Lussem, B.; Leo, K.
̈
Phys. Rev. B 2012, 85, 245209.
(21) Wunsche, J.; Reineke, S.; Lussem, B.; Leo, K. Phys. Rev. B 2010,
̈
̈
81, 245201.
(22) Hofmann, S.; Furno, M.; Lussem, B.; Leo, K.; Gather, M. C.
̈
Phys. Status Solidi A 2013, 210, 1467.
(23) Liehm, P.; Murawski, C.; Furno, M.; Lussem, B.; Leo, K.;
̈
Gather, M. C. Appl. Phys. Lett. 2012, 101, 253304.
(24) Scholz, R.; Gisslen
M.; Louis, E.; San Fabian
A 2009, 113, 315.
́
, L.; Himcinschi, C.; Vragovic,
Maroto, E.; Díaz-García, M. A. J. Phys. Chem.
́
I.; Calzado, E.
́
(25) Scholz, R.; Hofmann S.; et al., in preparation.
(26) Wu, C.-H.; Chien, C.-H.; Hsu, F.-M.; Shih, P.-I.; Shu, C.-F. J.
Mater. Chem. 2009, 19, 1464.
(27) Pabst, M.; Sundholm, D.; Koehn, A. J. Phys. Chem. A 2012, 116,
15203.
(28) Weller, A. Z. Z. Phys. Chem. (Muenchen, Ger.) 1982, 133, 93.
(29) Scholz, R.; Luschtinetz, R.; Seifert, G.; Jageler-Hoheisel, T.;
̈
Korner, C.; Leo, K.; Rapacioli, M. J. Phys. Condens. Matter 2013, 25,
473201.
̈
(30) Duhm, S.; Heimel, G.; Salzmann, I.; Glowatzki, H.; Johnson, R.
L.; Vollmer, A.; Rabe, J. P.; Koch, N. Nat. Mater. 2008, 7, 326.
(31) Vonlanthen, D.; Rotzler, J.; Neuburger, M.; Mayor, M. Eur. J.
Org. Chem. 2010, 2010, 120.
(32) Baldo, M. A.; Adachi, C.; Forrest, S. R. Phys. Rev. B 2000, 62,
10967.
(33) Jankus, V.; Winscom, C.; Monkman, A. P. J. Phys.: Condens.
Matter 2010, 22, 185802.
(34) Loudon, R.: The Quantum Theory of Light; Clarendon, Oxford,
1983.
(35) Wei, W.; Djurovich, P. I.; Thompson, M. E. Chem. Mater. 2010,
22, 1724.
(36) Bruker. SADABS User Manual - Area Detector Absorption and
Other Corrections V2.03; Delft, Netherlands, 2002.
(37) Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G. L.;
Giacovazzo, C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.;
Spagna, R. J. Appl. Crystallogr. 1999, 32, 115.
(38) Sheldrick, G. M. SHELXL-97 Program for the Refinement of
Crystal Structures; Universitat Gottingen: Gottingen (Germany), 1997.
̈
̈
̈
(39) Flack, H. D. Acta Crystallogr. 1983, A39, 876.
(40) Dunitz, J. D.; Orgel, L. E.; Rich, A. Acta Crystallogr. 1956, 9, 373.
(41) Ryan, M. F.; Eyler, J. R.; Richardson, D. E. J. Am. Chem. Soc.
1992, 114, 8611.
(42) Palsson, L.-O.; Monkman, A. P. Adv. Mater. 2002, 14, 757.
̊
2426
dx.doi.org/10.1021/cm500602y | Chem. Mater. 2014, 26, 2414−2426