Fluorinated copper(I) carboxylates as advanced tunable p-dopants for organic light-emitting diodes

Article abstract of DOI:10.1002/adma.201303252

Volatile copper(I) benzoates with variable degrees of fluorination are used for p-doping of organic hole-transport layers in single-carrier devices, charge-generation layers, and in organic light-emitting diodes. The charge-transport abilities of the dope

Full text of DOI:10.1002/adma.201303252

www.advmat.de
www.MaterialsViews.com
Fluorinated Copper(I) Carboxylates as Advanced Tunable
p-Dopants for Organic Light-Emitting Diodes
Günter Schmid,* Jan Hauke Wemken, Anna Maltenberger, Carola Diez, Arndt Jaeger,
Thomas Dobbertin, Oleksandr Hietsoi, Cristina Dubceac, and Marina A. Petrukhina*
Ever since Tang[1] et al. reported in 1987 the small-molecule
approach and later on Burroughes[2] et al. reported the polymer
approach to organic electroluminescence, scientists have been
dreaming of cheap, thin, lightweight, and flexible lighting tiles
and even wallpaper. More than 20 years after those publica-
tions the first pilot products appeared on the market,[3] not ful-
filling by far all the envisaged properties, yet. However, despite
tremendous improvements in lifetime, there are still major
challenges associated with product efficacy, cost, and off-state
appearance. One of the main goals involves the development
of a low-cost hole-transporting layer (HTL). As the closest vis-
ible layer in the final product, it is a major cost driver due to its
thickness and placement. Strong Lewis acids such as tungsten
or molybdenum oxides are used as very popular p-dopants.[4]
However their d-d splitting and strong interaction with the
host material introduces absorption and, thus, color into the
HTL (see photographs in Figure S1, Supporting Information).
Alternative organic dopants, such as 2,3,5,6-tetrafluoro-7,7,8,8-
tetracyano-quinodimethane (F4-TCNQ) and its derivatives, also
are strongly colored.[5] Therefore, better alternatives are greatly
needed to avoid unwanted coloration but keep all other charac-
teristics of an HTL. In this study, we propose to evaluate vola-
tile and electrophilic fluorinated copper(I) carboxylates as a new
class of p-dopants for HTL.
similar with processing an HTL in a small-molecule organic
light-emitting diode (OLED). Based on spectroscopic, single-
crystal diffraction, and complementary density functional theory
(DFT) studies, it was concluded that the materials are formed
by co-crystallization and not by complex formation, thereby
preserving the colorless appearance of the bulk crystals. These
properties, along with the ease of fine-tuning the Lewis acidity
of copper(I) carboxylate products, made this class of compounds
very promising candidates for this investigation. Therefore,
in this work we combined the efforts of chemists with mate-
rial and device manufacturers to test the new copper(I) dopants
aiming at the development of a low-cost HTL. From the device
perspective, the question arose if the donor-acceptor interac-
tions in these systems are strong enough to alter the conduc-
tivity of organic hole-transport materials. An increase of a factor
of 100–1000 is sufficient to lower the voltage drop in an OLED
over the HTL to some tens of millivolts, which is a necessary
prerequisite for high luminous efficacies. Herein, we found
that HTLs made of 2,2′,7,7′-tetra(N,N′-ditolyl)amino-9,9′-spiro-
bifluorene (Spiro-TTB) doped by copper(I) pentafluorobenzoate
(CupFBz) yield an OLED that fulfils device requirements.
In the following sections, we present a detailed study of
the electrical properties of the hole-transport layers doped by
the selected copper(I) carboxylate. Single-carrier devices were
used to study the increase in conductivity of hole transporters
depending on the copper(I) dopant. For the next step, we fur-
ther varied the structure of the hole transporter and measured
the conductivity with respect to the doping concentration. Addi-
tionally, the spectrally dependent extinction coefficient of the
developed doped hole-transporting material was determined
and compared to a similar system comprising the known p-type
dopant used as a reference, namely molybdenum oxide (MoO3).
After testing the whole series of fluorinated copper(I) benzo-
ates, CupFBz was selected due to its superior performance
compared to other members. Next, CupFBz in 4,4′,4″-tris(N-
(1-naphthyl)-N-phenylamino)triphenylamine (1-TNATA) was
used for further evaluation in charge-generation layers (CGL)
as interconnection units in stacked OLEDs to get white light-
emission.[9] It is revealed that both the electrical conductivity
and the optical transparency of the doped hole-transporting
material are excellent, especially when high HTL thicknesses
are required. Finally, the superior performance of the developed
p-type dopant in comparison to molybdenum oxide is shown in
a high efficiency stacked white OLED pilot prototype.
In the last several years, the Petrukhina group has prepared
a number of new fluorinated copper(I) carboxylates followed by
comprehensive studies of their structure-property correlations
with respect to the carboxylate ligand selected.[6] Interactions
of the resulting electrophilic polynuclear copper(I) complexes
with polyaromatic donors, such as coronene,[7] pyrene,[7b] coran-
nulene,[8] and others[7b] have also been examined. Notably, the
adducts were prepared by co-deposition of copper(I) carboxy-
lates with polyarenes at low pressure, a technique that is very
Dr. G. Schmid, J. H. Wemken, A. Maltenberger
Siemens AG, Corporate Technology
Guenther-Scharowsky-Strasse 1, 91058,
Erlangen, Germany
E-mail: guenter.schmid@siemens.com
Dr. C. Diez, Dr. A. Jaeger, Dr. T. Dobbertin
OSRAM Opto Semiconductors GmbH
Leibnizstrasse 4, 93055, Regensburg, Germany
Dr. O. Hietsoi, C. Dubceac, Prof. M. A. Petrukhina
Department of Chemistry
University at Albany, State University of New York
Albany, NY, 12222, USA
E-mail: mpetrukhina@albany.edu
Single-Carrier Devices: Copper(I) carboxylates were synthe-
sized according to published procedures starting from copper(I)
trifluoroacetate which is readily prepared by reacting copper(I)
oxide with a mixture of trifluoroacetic acid and its anhydride in
DOI: 10.1002/adma.201303252
©
878 wileyonlinelibrary.com
2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Mater. 2014, 26, 878–885

www.advmat.de
www.MaterialsViews.com
strictly anhydrous benzene or toluene.[10] Copper(I) trifluoro-
acetate has also been tested as a p-dopant. However, due to its
low sublimation temperature of 100–110 °C, it was difficult
to control the doping level in the p-conducting polyaromatic
host materials. Therefore, we moved to the higher molecular
weight congeners, a series of fluorinated copper(I) benzoates,
which were prepared by ligand exchange of copper(I) trifluroac-
etate with the corresponding commercially available carboxylic
acids.[6a,7a] Notably, sublimation temperatures of fluorinated
copper(I) benzoates reached the range of 180–200 °C that is
required for stable processing and good control of doping
level. The single-carrier devices were prepared by thermal
evaporation of fluorinated copper(I) benzoates doped into the
triarylamine hole conductor, namely N,N′-bis(naphthalen-1-yl)-
N,N′-bis(phenyl)-benzidine (α-NPD). Notably, α-NPD has a low
HOMO level of −5.5 eV[11] and therefore its ionization with a
weak dopant should not be expected. If, however, a doping effect
could be observed in α-NPD, it is very likely that other matrix
materials with higher HOMO levels can be readily ionized and
ought to show stronger doping effects.
We found that the selected copper(I) complexes exhibit a
Lewis acidity sufficient for doping purposes. Figure 1 shows the
influence of the degree and position of ligand fluorination of
the above copper(I) dopants on the current density.
Although copper(I) benzoate (circles in Figure 1a) exhibits
some influence on the current density, only the use of fluori-
nated copper(I) benzoates increases the current density over
the necessary level. Herein, the fluorination of the meta-posi-
tion on the benzoic ring (3-F) is favoured over the para-position
(4-F) (Figure 1a) and increases the current density alongside
with a shift towards lower voltages within the current rising
region (0.5 to 1.5 V). Raising the level of fluorination in the
meta-position (3,5-F) further increases the current density,
while increased fluorination in the ortho-position (2,6-F) lowers
the current density even below the undoped α-NPD reference
(Figure 1b). A further increase of fluorination in the meta- and
para-positions (3,4,5-F) combines the efficient shift in voltage
of the meta-positions (3,5-F) with the decrease in current den-
sity of the para-position (4-F). The fully fluorinated copper(I)
pentafluorobenzoate yields the highest current density and the
strongest shift in voltage (−0.8 V) of the rising region. It can
be related with the fact that increasing level of ligand fluori-
nation raises the Lewis acidity of the copper centers, thereby
enhancing the strength of the donor-acceptor interactions with
the host molecules. This gradual increase of doping strength
with higher ligand fluorination degree was confirmed with a
MOPAC PM6 simulation. Herein, the relative LUMO energy
level of copper(I) benzoates gradually decreases by a total of
1 eV with increasing number of F-substituent on carboxylate
ligands from monofluorinated (dropping from meta- to para-)
to pentafluorinated ones. The resulting graphical display of this
decrease is shown in Figure 2.
Figure 1. Electrical data of single-carrier devices of copper(I) benzoates
with increasing degree of fluorination doped with vol. 15% into α-NPD:
a) Unfluorinated and mono-fluorinated copper(I) benzoates compared
to undoped α-NPD, b) Bis-, tris- and pentafluoro copper(I) benzoates
compared to undoped α-NPD. n-F indicates the position of F-atoms on
the benzoic ring.
by the generation of free carriers resulting in an increased
conductivity.
To determine the influence of the host’s HOMO level on
the doping performance of copper(I) pentafluorobenzoate, we
investigated the well-known triarylamine-based hole conduc-
tors, namely 1-TNATA and Spiro-TTB. Single-carrier devices
were prepared and the measured conductivities are displayed
in Figure 3a. For reference, the data of the well-known hole-
transporting material α-NPD:MoO3 is also shown.
The ideal coordination and thereby partial charge transfer
occurs when HOMO level of the host is approximately matching
or higher than LUMO level of the dopant. Spiro-TTB and
1-TNATA feature a HOMO level of −5.3 eV[13] and −5.0 eV,[15]
respectively, which fits to the LUMO level of the dopant. For
α-NPD (HOMO = −5.5 eV)[11,14] this prerequisite is not ful-
filled. The resulting conductivity in the host Spiro-TTB was
higher compared to 1-TNATA due to its higher mobility.
The calculated decrease in LUMO energy levels enhances
the possibility of a charge transfer from the HOMO level of
the host to the LUMO level of the dopant, thereby increasing
the number of generated free charge carriers. The resulting
conductivity is proportional to the product carrier concentra-
tion and mobility within the host material. Adding dopants to
a host generally reduces the mobility which is overcompensated
©
wileyonlinelibrary.com
Adv. Mater. 2014, 26, 878–885
2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
879

www.advmat.de
www.MaterialsViews.com
of-the-art dopants such as MoO3 is really advantageous, espe-
cially for the use of thick hole-transport layers as demonstrated
in detail later. For our OLED work, we finally choose Spiro-TTB
over 1-TNATA as host material for CupFBz due to its deeper
HOMO, while retaining low absorption and sufficiently high
conductivity.
Regarding the doping mechanism itself, two ways of inter-
action of the dopant and host materials are possible. First, the
nitrogen atoms of the hole conductor materials can interact
with the electrophilic centers of the fluorinated copper(I) car-
boxylate in a coordination manner, thereby, formally leaving a
hole on the host. This hole can be delocalized on the molecule
or can even hop along the percolated hole transporter in the
co-evaporated amorphous solid layer. This mechanism is sche-
matically illustrated in Figure 4. For simplicity only a triphe-
nylamine is shown.
The second possibility includes Cu•••π interactions observed
in a number of electrophilic polynuclear copper(I) complexes
with planar and curved polyarenes.[7,8] Regardless of the type,
both interaction modes induce a partial electron transfer from
host to dopant and subsequently enhance the conductivity of
the percolated host material.
Figure 2. Simulated relative HOMO/LUMO levels for a series of copper(I)
benzoates. LUMO energy levels of the gradually increasing fluorination in
the compounds are displayed relative to 2-F set. HOMO energy levels are
displayed relative to 2-F set at −3 eV for visual reasons.[12] n-F indicates
the position of F-atoms on the benzoic ring. The lines in the graph are for
orientation of the relative changes.
The observed efficient function of CupFBz as a p-dopant in
various host materials forms the basis for its further exploita-
tion in charge-generation layers and OLEDs as shown in the
following sections.
Higher conductivities such as with Spiro-TTB:MoO3 are pos-
sible, but such strong dopants usually show a high absorption
and high extinction coefficient.[4b] However, this is not seen for
Spiro-TTB:CupFBz, as shown in Figure 3b. The observed low
extinction coefficient in the visible range compared to the state-
Charge-Generation Layers: Charge-generation layers (CGL)
are key elements between the emitting units in stacked
OLEDs.[9c,11] In recent years, stacked OLEDs have drawn con-
siderable attention due to their superior luminous efficiency
and brightness in conjunction with long operational lifetimes
compared to conventional OLEDs with only a single emitting
unit. The stacking approach reduces the operational stress
for each emitting unit, resulting in enhanced OLED lifetime
at a given luminance. The emitting units are connected by
CGLs containing a p-n heterojunction.[16] In this case the field-
induced carrier generation occurs at the doped organic/organic
heterointerface via tunnelling of electrons through the deple-
tion zone from the HOMO of the p-type layer to the LUMO of
the n-type layer.
For evaluating the working mechanism of the developed p-
type dopant in a charge-generation layer a simplified test struc-
ture was developed. The layer sequence is shown in the inset
of Figure 5a. The test structure is designed so that by applying
a positive bias to the indium tin oxide (ITO) anode a tunnel-
ling current can be measured. The undoped 1-TNATA and
NOVALED electron conductor NET-18[17] layers prevent direct
charge carrier injection from the electrodes to ensure that the
measured current results from charge separation at the doped
p-n junction.
Figure 5a shows the current density-voltage characteristic of
the processed CGL test structure with CupFBz as a p-type dopant
in the hole-transporting material, 1-TNATA. The n-type dopant
is the commercially available NOVALED n-dopant NDN-26 in
the electron transporting host NET-18.[17,18] To prevent chemical
reactions at the interface between the opposing doped layers a
4 nm-thick copper phthalocyanine (CuPc) interlayer is found to
considerably improve the long-term stability of doped CGLs.[16b]
In Figure 5b, the time dependence of the voltage at a constant
Figure 3. Electrical and optical data of CupFBz in different hosts in com-
parison to state-of-the-art p-type dopants. Concentrations are given in
vol%. a) Conductivity with respect to doping concentration. b) Refractive
index n and extinction coefficient k in the visible and adjacent ranges.
©
880 wileyonlinelibrary.com
2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Mater. 2014, 26, 878–885

www.advmat.de
www.MaterialsViews.com
Figure 4. Interaction of copper(I) pentafluorobenzoate and the host material: a) Coordination of the host molecule to the copper core of the dopant;
b) partial charge transfer, generation of a free delocalized hole on the host molecule, and charge transport via hopping. For the chemical structures of
Spiro-TTB, 1-TNATA, and α-NPD, see experimental section.
current density of 10 mA/cm2 is displayed. Note that this cur-
rent density corresponds to accelerated testing conditions
in OLEDs. Within the first hours, the measured data show a
voltage drop resulting from the temperature enhanced conduc-
tivity of organic materials due to heating during the measure-
ment.[19] After temperature equilibrium is reached the voltage
remains constant over time, which is highly desirable for the
application in a stacked OLED. Therefore, a CGL comprised of
CupFBz as a p-type dopant is suitable in state-of-the-art stacked
white OLEDs for lighting applications.
Organic Light-Emitting Diodes: For large-area production of
prototype OLEDs, sufficient hole-injection layer (HIL) thick-
nesses of several hundreds of nanometers are needed to cover
particles on the ITO substrate and to avoid short circuits.[20]
Therefore, it is required that both the electrical conductivity and
the optical transparency should be excellent. To demonstrate
the superior electrical and optical behaviour of the developed
p-type CupFBz dopant in a hole-transporting material, organic
light-emitting diodes with varying HIL thicknesses were mod-
elled, processed, and characterized.
The device structure of the fabricated monochrome OLEDs
using the phosphorescent material fac-tris(2-phenylpyridine)
iridium(III) (Ir(ppy)3) as an emitter in a predominantly elec-
tron- conductive host material (TMM-004)[21] is schematically
displayed in Figure 6a. The HIL thicknesses of both materials,
Spiro-TTB:CupFBz and α-NPD:MoO3, were varied between
30 nm and 420 nm. Both OLED series were processed on the
same ITO substrate within a single fabrication run using com-
binatorial shadow masks during evaporation. This processing
technique ensures entire comparability of the devices. The
measurement of the external quantum efficiency (EQE) was per-
formed in an integrating sphere without outcoupling enhance-
ment via half ball lenses, microlens arrays, or scattering foils.[22]
The substrate edges were blackened to avoid influences from
side emissions of the OLEDs. By using optical simulations
based on a dipole model[23] the experimental results of the meas-
ured external quantum efficiencies of both OLED series were
Figure 5. Electrical data of CGL test structure. a) Current density–voltage
characteristics. b) Long-term voltage stability at a constant current den-
sity of 10 mA/cm2.
©
wileyonlinelibrary.com
Adv. Mater. 2014, 26, 878–885
2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
881

www.advmat.de
www.MaterialsViews.com
same as shown in Figure 5a (50 nm 1-TNATA:CupFBz/4 nm
CuPc/ 50 nm NET18:NDN26). The HIL thickness of both large
area OLED devices (40 cm2) was 300 nm and the doping con-
centration of both p-type dopants was 5%. The current density–
voltage characteristics of the two devices were nearly identical
(Figure 7b), indicating that the electrical properties such as the
conductivity of these HIL materials are comparable and a min-
imum of 5% dopant concentration is sufficient for both dopants.
All the measurements were done with a scattering foil attached
to the ITO surface for external light outcoupling.[22] Figure 7c
shows the electroluminescence spectra measured at a constant
current density of 3 mA/cm2 resulting in comparable color coor-
dinates of CIE (0.44/0.40) for the device with CupFBz and CIE
(0.45/0.39) for the device with MoO3, respectively. It is striking
that the radiance of the CupFBz device is higher compared to
the MoO3 device. This means that the former is more efficient
compared to the latter. The efficiency measurement confirms
this assumption and yields a current efficiency and luminous
efficacy of 58 cd/A and 27 lm/W, respectively, at a common
luminance of 1000 cd/m2 for the device with MoO3 and 59 cd/A
and 32 lm/W, respectively, (at 1000 cd/m2) for the device with
CupFBz. This efficiency enhancement can be attributed to
the very low absorption properties of the device with CupFBz
doped Spiro-TTB HIL. By attaching the scattering foil for out-
coupling enhancement an additional interface is introduced
resulting in multiple light propagation within the organic stack
and, therefore, enhanced absorption probability in the doped
HIL. Thus, organic materials with a very low extinction coeffi-
cient and hence absorption are in this case even more important
compared to devices without outcoupling enhancement. These
results clearly demonstrate that it is possible to fabricate state-
of-the-art large area stacked white organic light-emitting diodes
using the developed cost efficient p-type dopant CupFBz.
Herein, we demonstrated for the first time that fluorinated
copper(I) benzoates, especially the perfluorinated compound,
have sufficient Lewis acidity to promote the hole transport in tri-
arylamine-based host materials. Importantly, we also found that
they fulfil simultaneously all required properties for OLED fab-
rication: i) the partial charge transfer between the dopant and
the host is sufficiently large to enhance conductivity, but ii) low
enough not to enhance the absorption of such HTLs as the off-
state appearance of devices based on CupFBz doped HIL still
remains neutral in contrast to many organic low electron density
compounds such as F4-TCNQ and its higher molecular weight
derivatives and inorganic materials such as MoO3, respectively;
iii) the synthesis uses readily available starting materials and
relies on a simple and effective route, thus providing a base for
very low cost dopants; iv) the processing parameters fit well into
the temperature range for processing small-molecule OLEDs;
v) we tested the CupFBz doping within our current product
ORBEOS white OLED stack which has an area of 49 cm2, a
luminous efficacy of 23 lm/W and a lifetime (luminance drop
down to 50% of initial value) of 5000 h at j = 3.8 mA/cm2 (corre-
sponding to I = 186 mA). Devices with CupFBz doped HIL were
processed and tested. The lifetime of such ORBEOS devices was
similar to or even exceeding that of standard ORBEOS white
OLED stack at comparable electro-optical performance. The
full specifications of ORBEOS are included in the Supporting
Information.[24]
Figure 6. Device structure and performance of OLEDs with varying HIL
thicknesses. a) Sketch of the device architecture. b) Calculated (solid line)
and measured (points) external quantum efficiencies with respect to HIL
thickness of two different HIL materials: CupFBz doped Spiro-TTB and
MoO3 doped α-NPD, respectively.
validated (Figure 6b). It is remarkable that the EQE of the OLEDs
with CupFBz doped Spiro-TTB is only modulated by optical
interference due to microcavity effects. This indicates that,
on one hand, the electrical conductivity of Spiro-TTB:CupFBz
(5%) is sufficient to avoid an additional voltage drop across the
hole injection layer. On the other hand, no complex formation
resulting in absorption in the visible range is observed.
For devices with the conventional HIL, α-NPD:MoO3 mate-
rial, the optical microcavity effect is overlaid by a reduction in
EQE with increasing HIL thickness. This can be attributed to
the high extinction coefficient between 400 nm and 550 nm
(Figure 3b) resulting in absorption especially at large HIL thick-
nesses. We have to point out that these findings demonstrate
again the superior working mechanism of the developed p-type
dopant, namely CupFBz.
Finally, a high efficient stacked OLED production prototype
using the CupFBz dopant in both the HIL and CGL was pro-
cessed and characterized. For comparison, a similar OLED
comprising molybdenum oxide as a p-type dopant was fab-
ricated. The device comprised of the stacked white OLEDs is
shown in Figure 7 along with the electro-optical data.
As schematically shown in Figure 7a, the electron injec-
tion layer (EIL), the emission units (EMUs), and the CGL of
both OLED devices were essentially identical except for the
p-type dopant. The EIL consists of the proprietary material set
from Novaled AG[17] NET18:NDN26 (8%) and the CGL is the
©
882 wileyonlinelibrary.com
2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Mater. 2014, 26, 878–885

www.advmat.de
www.MaterialsViews.com
Figure 7. Electro-optical characterization of a cost-efficient state-of-the-art stacked OLED pilot prototype with CupFBz in comparison to MoO3 as p-
type dopants. a) Simplified sketch of the device architecture. b) Current density–voltage characteristics. c) Electroluminescence spectra measured at a
constant current density of 3 mA/cm2. d) Current efficiency and luminous efficacy with respect to luminance.
Commercial OLED materials were purchased from NOVALED[17]
Experimental Section
(NET-18, NDN-26) and Merck (TMM-004, Ir(ppy)3).[21] NOVALED and
Merck do not disclose the chemical structure of their high-performance
materials. Advice for inclusion into OLED-devices can be obtained from
these suppliers. MoO3 was obtained from Sigma–Aldrich. All other OLED
materials were ordered from LUMTEC Corp. and used without further
purification.
The chemical structures of the hole conductors Spiro-TTB, 1-TNATA,
and α-NPD used as host materials for the fluorinated copper(I)
carboxylates are depicted in Figure 8. The molecular structure of the
copper(I) carboxylate is shown in Figure 4.
Single-carrier devices were fabricated by thermal vapor deposition
in a high vacuum system within an argon filled glovebox. ITO-covered
glass substrates were employed and the ITO was oxygen-plasma
activated for 10 min before layer deposition. The doped layers of 200 nm
thickness and constant concentrations of 15 vol% in α-NPD for all
copper(I) complexes were sandwiched between the ITO anode and an
evaporated 150 nm-thick aluminium cathode. Shadow masks were used
to create separate size defined devices on the same substrate. Voltage-
driven current (IV)-measurements were performed straight after device
fabrication inside an argon-filled glovebox using a computer-controlled
Keithley 237 voltage source.
Experimental manipulations involving the synthesis of anhydrous copper(I)
carboxylates were carried out using standard Schlenk and glove-box
techniques. All the carboxylic acids were used as received from Oakwood
Chemical. Benzene was dried over sodium/benzophenone in an argon
atmosphere and freshly distilled prior to use. Toluene was dried over sodium
in an argon atmosphere and freshly distilled prior to use. Several copper(I)
benzoates were synthesized by the previously reported ligand exchange
method starting from copper(I) trifluoroacetate[10] and the respective
carboxylic acids: 3-fluoro-,[6a] pentafluoro-,[6a] 3,5-difluoro-,[7a] and benzoic
acid.[25] The details on preparation of copper(I) 4-fluoro-, 2,6-difluoro-, and
3,4,5-trifluorobenzoates are reported below.
Copper(I) 3,4,5-trifluorobenzoate was prepared by an overnight reflux
of Cu4(O2CCF3)4 (0.642 g, 0.9 mmol) and (3,4,5-F)3C6H2COOH (1.160 g,
6.59 mmol) in benzene (60 mL). The solvent was then removed under
vacuum to afford a very pale blue powder. The product was further
purified under dynamic vacuum to remove the traces of unreacted
acid. This afforded a light microcrystalline powder with sublimation
temperature of 185–200 °C. Yield: 0.580 g (66.5%).
Copper(I) 4-fluorobenzoate and copper(I) 2,6-difluorobenzoate were
prepared by similar procedures.
©
wileyonlinelibrary.com
Adv. Mater. 2014, 26, 878–885
2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
883

www.advmat.de
www.MaterialsViews.com
Figure 8. Chemical structures of the hole conductor materials Spiro-TTB, 1-TNATA, and α-NPD.
All single-carrier and OLED devices prepared in this study exhibited
Acknowledgements
an active area of 4 mm[2] and were fabricated in a vacuum-deposition
chamber at a base pressure of 10−7 mbar. Before evaporation, the
patterned ITO glass substrates were cleaned with solvents in a
multistep process and exposed to oxygen plasma. All the organic
materials were used as received. The copper(I) carboxylates were triply
sublimed. The rate of deposition for the organic materials was in total
0.05 nm/s, and the substrates were not heated during evaporation. The
doped layers were directly formed from the gas phase by coevaporation
of the host material and the dopant at rates in the ratio corresponding
to the volumetric doping concentration (vol%). This deposition
technique was also used for the red, green, and blue dye inside the
host. Prior to testing, the devices were encapsulated under inert gas
atmosphere with a glass lid and getter containing zeolite as a dryer.
Current density–voltage characteristics and electroluminescence (EL)
spectra were measured using a source measurement unit (Keithley
2400-C) and a spectrometer (Instrument Systems CAS 140 with
TOP100).
The authors (G.S., J.H.W., A.M., C.D., A.J., T.D.) would like to thank the
German Federal Ministry of Education and Research (BMBF) for funding
part of this work under contract FKZ 13N10474 (TOPAS 2012) and FKZ
13N22240 (OLYMP). Partial support of this work from the U.S. National
Science Foundation (CHE-1212441) is gratefully acknowledged by M.A.P.
Received: July 15, 2013
Revised: September 4, 2013
Published online: December 20, 2013
[1] C. W. Tang, S. A. VanSlyke, Appl. Phys. Lett. 1987, 51, 913.
[2] J. H. Burroughes, D. D. C. Bradley, A. R. Brown, R. N. Marks,
K. Mackay, R. H. Friend, P. L. Burns, A. B. Holmes, Nature 1990,
347, 539.
MOPAC (Molecular Orbital Package) is a semi-empirical quantum
chemistry program based on Dewar and Thiel’s NDDO approximation.[26]
In this work, MOPAC2009 was used with its latest simulation method
PM6. A simple molecular structure based on the known structure of
copper(I) 3-fluorobenzoate[6a] was used to create the separate input files
(see Supporting Information for details). Herein, the core structure of
all the simulated compounds was kept constant and only the ligand
fluorination was modified by a stepwise replacement of hydrogen with
fluorine. The corresponding distances were modified from C–H (approx.
0.95 Å) to C–F (approx. 1.37 Å) accordingly.
[3] a) B. D‘Andrade, Nat. Photonics 2007, 1, 33; b) S. Logothetidis,
Mater. Sci. Eng., B 2008, 152, 96; c) NanoMarkets, OLED Lighting
Materials Market Forecasts, 2013, Report #Nano-577.
[4] a) S. Hamwi, J. Meyer, T. Winkler, T. Riedl, W. Kowalsky, Appl.
Phys. Lett. 2009, 94, 253307/1; b) J.-H. Lee, D.-S. Leem, H.-J. Kim,
J.-J. Kim, Appl. Phys. Lett. 2009, 94, 123306/1; c) M.-T. Hsieh,
C.-C. Chang, J.-F. Chen, Appl. Phys. Lett. 2006, 89, 103510/1;
d) C.-C. Chang, M.-T. Hsieh, J.-F. Chen, S.-W. Hwang, C. H. Chen,
Appl. Phys. Lett. 2006, 89, 253504/1.
[5] a) C.-H. Chang, H.-C. Cheng, Y.-J. Lu, K.-C. Tien, H.-W. Lin, C.-L. Lin,
C.-J. Yang, C.-C. Wu, Org. Electron. 2010, 11, 247; b) L.-S. Liao,
K. P. Klubek, C. W. Tang (Eastman Kodak Company), US Patent
7,528,545, 2009.
[6] a) Y. Sevryugina, O. Hietsoi, M. A. Petrukhina, Chem. Commun.
2007, 3853; b) Y. Sevryugina, D. D. Vaughn II, M. A. Petrukhina,
Inorg. Chim. Acta 2007, 360, 3103.
Supporting Information
Supporting Information is available from the Wiley Online Library or
from the author.
©
884 wileyonlinelibrary.com
2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Mater. 2014, 26, 878–885

www.advmat.de
www.MaterialsViews.com
[7] a) Y. Sevryugina, A. Yu. Rogachev, M. A. Petrukhina, Inorg. Chem.
2007, 46, 7870; b) Y. Sevryugina, M. A. Petrukhina, Eur. J. Inorg.
Chem. 2008, 219.
[8] Y. Sevryugina, E. A. Jackson, L. T. Scott, M. A. Petrukhina, Inorg.
Chim. Acta 2008, 361, 3103.
[18] a) J. Birnstock, T. Canzler, M. Hoffman, A. Lux, S. Murano,
P. Wellmann, A. Werner, J. Soc. Inf. Displ. 2008, 16, 221;
b) J. Blochwitz-Nimoth, O. Langguth, S. Murano, G. He,
T. Romainczyk, J. Birnstock, J. Soc. Inf. Displ. 2010, 18, 595.
[19] a) Z. Shen, P. E. Burrows, V. Bulovic, D. M. McCarty,
M. E. Thompson, S. R. Forrest, Jpn. J. Appl. Phys., Part 2 1996,
35, L401; b) P. E. Burrows, Z. Shen, V. Bulovic, D. M. McCarty,
S. R. Forrest, J. A. Cronin, M. E. Thompson, J. Appl. Phys. 1996,
79, 7991; c) S. R. Forrest, Z. Shen, P. E. Burrows, V. Bulovic,
M. E. Thompson, in Inorganic and Organic Electroluminescence:
International Workshop on Electroluminescence, (Eds: R. H. Mauch,
H.-E. Gumlich), Berlin, Germany 1996, 83.
[20] M. Fehrer, K. Heuser, E. Hölfling, T. Schlenker (OSRAM Opto Semi-
conductors), US Patent 2010/0314648 A1, 2010.
[21] TMM-004 is a commercial material form Merck KGaA, Darmstadt,
Germany, see http://www.merck-performance-materials.de/de/
display/oled_materials/oled_materials.html, accessed: December
2013. Product data are available from the vendor.
[22] D. S. Setz, T. D. Schmidt, M. Flämmich, S. Nowy, J. Frischeisen,
B. C. Krummacher, T. Dobbertin, K. Heuser, D. Michaelis,
N. Danz, W. Brütting, A. Winnacker, J. Photonics Energy 2011, 1,
011006/1.
[23] a) R. R. Chance, A. Prock, R. Silbey, Adv. Chem. Phys. 1978, 37, 1;
b) N. Danz, R. Waldhäusl, A. Bräuer, R. Kowarschik, J. Opt. Soc.
Am. B 2002, 19, 412; c) B. C. Krummacher, S. Nowy, J. Frischeisen,
M. Klein, W. Brütting, Org. Electron. 2009, 10, 478.
[9] a) Z. Shen, P. E. Burrows, V. Bulovic, S. R. Forrest, M. E. Thompson,
Science 1997, 276, 2009; b) G. Gu, G. Parthasarathy, P. Tian,
P. E. Burrows, S. R. Forrest, J. Appl. Phys. 1999, 86, 4076; c) J. Kido,
T. Matsumoto, T. Nakada, J. Endo, K. Mori, N. Kawamura, A. Yokoi,
SID Int. Symp. Digest Tech. Papers 2003, 34, 979; d) Q. Y. Bao,
J. P. Yang, Y. Q. Li, J. X. Tang, Appl. Phys. Lett. 2010, 97, 063303/1;
e) L.-S. Liao, W. K. Slusarek, T. K. Hatwar, M. L. Ricks, D. L. Comfort,
Adv. Mater. 2008, 20, 324; f) T. Chiba, Y.-J. Pu, R. Miyazaki,
K. Nakayama, H. Sasabe, J. Kido, Org. Electron. 2011, 12, 710.
[10] F. A. Cotton, E. V. Dikarev, M. A. Petrukhina, Inorg. Chem. 2000, 39, 6072.
[11] H. Kanno, R. J. Holmes, Y. Sun, S. Kena-Cohen, S. R. Forrest, Adv.
Mater. 2006, 18, 339.
[12] The simulated HOMO/LUMO energy levels can only be accounted
relative to each other. The absolute values are not corresponding
to the real values due to the limits of the simulation method. The
relative decrease in LUMO energy with increasing fluorine concen-
tration, however, represents the increasing doping efficiency with
increasing fluorine concentration.
[13] K. Walzer, B. Maennig, M. Pfeiffer, K. Leo, Chem. Rev. 2007, 107, 1233.
[14] J. Staudigel, M. Stößel, F. Steuber, J. Simmerer, J. Appl. Phys. 1999,
86, 3895.
[15] J. Staudigel, M. Stößel, F. Steuber, J. Blässing, J. Simmerer, Synth.
Met. 2000, 111–112, 69.
[16] a) L. S. Liao, K. P. Klubek, C. W. Tang, Appl. Phys. Lett. 2004, 84,
167; b) C. Diez, T. C. G. Reusch, E. Lang, T. Dobbertin, W. Brütting,
J. Appl. Phys. 2012, 111, 103107/1.
[17] NET-18, NDN-26 are commercial materials from Novaled AG,
Dresden, Germany. http://www.novaled.com/products/organic_
material/for_oled_lighting/, accessed: December 2013. Product
data are available from the vendor.
[24] ORBEOS reference data can be downloaded from the web page
of our distributor: http://docs-europe.electrocomponents.com/
webdocs/0d96/0900766b80d967e8.pdf, accessed: December 2013.
The full dataset is included into the Supporting Information.
[25] D. A. Edwards, R. Richards, J. Chem. Soc., Dalton Trans. 1973,
2463.
[26] James J. P. Stewart, MOPAC2009, Stewart Computational Chem-
istry, Colorado Springs, CO, USA, http://openmopac.net, accessed:
September 2013.
©
wileyonlinelibrary.com
Adv. Mater. 2014, 26, 878–885
2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
885