Creating a pseudometallic state of K+ by intercalation into 18-crown-6 grafted on polyfluorene as electron injection layer for high performance PLEDs with oxygen- and moisture-stable Al cathode

Article abstract of DOI:10.1021/ja111521p

Polymer light-emitting diodes (PLEDs) suffer from inadequate lifetimes because of the use of environmentally sensitive metals as the cathodes. We present the use of water/methanol-soluble polyfluorene grafted with 18-crown-6 chelating to K⁺ as the electron-injection layer (EIL) for deep-blue-emission PLEDs, allowing the use of environmentally stable Al as the cathode since electron donation from the 18-crown-6 can reduce K⁺ to a stable "pseudometallic state", enabling it to act as an intermediate step for electron injection. Furthermore, when poly(ethylene oxide) was blended into the EIL to provide hole blocking (HB), the device exhibited the highest performance reported to date for a deep-blue-emission PLED based on a conjugated polymer as the emitting layer, with a brightness of 54-800 cd/m² and an external quantum efficiency of 5.42%. The use of such an EI-HB layer opens a broad avenue leading toward industrialization of PLEDs.

Full text of DOI:10.1021/ja111521p

COMMUNICATION
pubs.acs.org/JACS
Creating a Pseudometallic State of Kþ by Intercalation into 18-Crown-
6 Grafted on Polyfluorene as Electron Injection Layer for High
Performance PLEDs with Oxygen- and Moisture-Stable Al Cathode
Hsin-Hung Lu, Yun-Sheng Ma, Neng-Jye Yang, Guan-Hong Lin, Yun-Chung Wu, and Show-An Chen*
Chemical Engineering Department, National Tsing-Hua University, Hsinchu 30041, Taiwan, ROC
S Supporting Information
b
poly(9,9-dihexylfluorene)-based device with Ca as the cathode to
ABSTRACT: Polymer light-emitting diodes (PLEDs) suf-
fer from inadequate lifetimes because of the use of environ-
mentally sensitive metals as the cathodes. We present the
use of water/methanol-soluble polyfluorene grafted with
18-crown-6 chelating to Kþ as the electron-injection layer
(EIL) for deep-blue-emission PLEDs, allowing the use of
environmentally stable Al as the cathode since electron
donation from the 18-crown-6 can reduce Kþ to a stable
“pseudometallic state”, enabling it to act as an intermediate
step for electron injection. Furthermore, when poly-
(ethylene oxide) was blended into the EIL to provide hole
blocking (HB), the device exhibited the highest perfor-
mance reported to date for a deep-blue-emission PLED
based on a conjugated polymer as the emitting layer, with a
brightness of 54 800 cd/m2 and an external quantum
efficiency of 5.42%. The use of such an EIꢀHB layer opens
a broad avenue leading toward industrialization of PLEDs.
reduce the turn-on voltage Von from 6.6 V (without PFC)
to 4.1 V (with PFC) and enhance the maximum brightness Bmax
(and ηL) from 880 cd/m2 (0.29 cd/A) to 2800 cd/m2 (0.53 cd/A);
this was attributed to the formation of an interfacial dipole.8 Crown
ethers are a special class of ether able to form stable complexes with
alkali, alkaline-earth, and transition-metal ions.9,10 A metal ion with
diameter close to the cavity diameter of a crown ether can form a
stable complex with it.10 For example, the cavities of 12-crown-4
(1.2ꢀ1.5 Å), 15-crown-5 (1.7ꢀ2.2 Å), and 18-crown-6 (2.6ꢀ3.2 Å)
can form stable complexes with Liþ (1.36 Å), Naþ (1.94 Å), and Kþ
(2.66 Å), respectively.10
Here we propose the use of a water- or alcohol-soluble
18-crown-6 (Cn6)-grafted polyfluorene (PCn6; see Chart S1 in
the Supporting Information) blended with K2CO3 as the EIL in a
PLED with β-phase-containing poly(9,9-di-n-octylfluorene)
(β-PFO) as the EML. β-PFO, which exhibits a deep-blue emission
with high spectral stability in electroluminescence (EL),11 here
was formed by spin-coating of ethyl acetate on top of an
amorphous PFO film. [It should be noted that the photolumines-
cence quantum efficiency (PLQE) of β-PFO is 62%, which is
higher than that of amorphous PFO (39%)]. The solubility in
highly polar solvents (e.g., water or alcohol) provided by Cn6 can
prevent dissolution of the EML having a thin EIL atop it, and the
intercalation of Kþ into Cn6 (Scheme 1) allows the Kþ ion to act
similarly to potassium metal (and here is termed a “pseudometallic
state”) in reducing the electron-injection barrier from a stable
metal cathode (e.g., Al or Au) and facilitating electron transport.
Furthermore, blending poly(ethylene oxide) (PEO) into the
K2CO3-blended PCn6 layer (at a Cn6:Kþ mole ratio of 1:3) to
provide hole-blocking (HB) functionality can remarkably enhance
the device performance, affording the highest performance re-
ported to date for a deep-blue PLED with a conjugated polymer as
the EML. Chemical structures of 15-crown-5 (Cn5)-grafted
polyfluorene (PCn5), PEO, and PFO are shown in Chart S1.
We first define the notation for the EIꢀHB layers used in this
study. The label PCn6:Kþ(1:x) indicates that the layer is
composed of PCn6 and K2CO3, and 1:x in the parentheses
(x = 0, 1, 3) is the Cn6:Kþ mole ratio. PCn6:PEO(1:y) indicates
that the layer is composed of PCn6, K2CO3, and PEO with a
Cn6:Kþ mole ratio of 1:3, and 1:y in parentheses (y = 0.75, 1, 1.5, 2)
is the PCn6:PEO mass ratio. Finally, PCn6:Kþ(1:x)/Al
[or PCn6:PEO(1:y)/Al] and PCn6:Kþ(1:x)/Au [or PCn6:
PEO(1:y)/Au] denote β-PFO-based devices with Al and Au,
he development of polymer light-emitting diodes (PLEDs)
T
with practically acceptable lifetimes is an important issue for
realization of their industrialization. Therefore, the use of en-
vironmentally stable high-work-function metals such as Al, Cu,
Ag, and Au as the cathode has attracted extensive attention
recently. Use of a water- or alcohol-soluble electron-injection layer
(EIL) based on a conjugated polymer grafted with amino, ammo-
nium salt, or diethanolamino groups has been demonstrated to
allow the use of a high-work-function metal as the cathode because
the formation of interfacial dipole or space charge between the
emitting layer (EML) and the cathode can reduce the electron-
injection barrier.1ꢀ7 Consequently, the maximum external quan-
tum efficiencies ηext (and the corresponding maximum luminous
efficiencies ηL) for blue-, green-, and red-emission PLEDs using
fluorescent conjugated polymers as the EMLs and Al as the cathode
have been reported to reach 1.62% (1.3 cd/A),2 7.85% (23.8 cd/A),2
and 2.94% (2.89 cd/A),4 respectively. However, the brightnesses
(at the applied voltage) at the ηext for the three emissions are only
380 cd/m2 (9.7 V), 7923 cd/m2 (8.8 V), and 1040 cd/m2
(9.4 V), respectively, which are far lower than the deep-blue-
emission brightness of 54 800 cd/m2 (6.6 V) at 5.42% (6.14 cd/A)
obtained in the present work, as revealed below.
In addition to the above hydrophilic groups,1ꢀ7 crown ether
groups may be expected to serve the same purpose. Polyfluorene
grafted with 15-crown-4 moieties (PFC) was used as an EIL for a
Received: December 22, 2010
Published: June 06, 2011
r
2011 American Chemical Society
9634
dx.doi.org/10.1021/ja111521p J. Am. Chem. Soc. 2011, 133, 9634–9637
|

Journal of the American Chemical Society
COMMUNICATION
Scheme 1. Chelation of Cn6 to Kþ
Table 1. Characteristic PLED Performance Values for β-
PFO-Based Devices without and with EIꢀHB Layersa
Von (V)b
Bmax (cd/m2)
ηL (cd/A)
without EIꢀHB layer
PCn6:Kþ(1:0)
PCn6:Kþ(1:1)
10
12
624
0.017
0.17
1.88
2.03
6.14
5.95
5.45
1.65
5.8
3.6
3.6
3.4
3.8
3.4
3.8
18300
26500
54800
47100
38800
8900
PCn6:Kþ(1:3)
PCn6:PEO(1:0.75)
PCn6:PEO(1:1)
PCn6:PEO(1:1.5)
PCn6:PEO(1:2)
a The device structure was ITO/PEDOT (25 nm)/β-PFO
(120 nm)/[without/with EIꢀHB layer (20 nm)]/Al (60 nm). ITO
and PEDOT are abbreviations for indium tin oxide and poly(styrene
sulfonic acid)-doped poly(3,4-ethylenedioxythiophene), respectively.
b Brightness at 2 cd/m2.
respectively, as the cathode. This notation system has also been
applied to the case of PCn5 blended with K2CO3 and to the cases
of PCn6 and PCn5 blended with Li2CO3, Na2CO3, and Cs2CO3
with a crown ether:metal ion mole ratio of 1:1. In addition, PEO:
Kþ(6:3) indicates that the layer contains PEO and K2CO3 at a
PEO (repeat unit):Kþ mole ratio of 6:3.
Table 1 shows the characteristic PLED performance of
β-PFO-based devices without and with EIꢀHB layers, and their
characteristic curves of current density (J) and brightness (B)
versus voltage and of luminous efficiency versus J are shown in
Figure 1. The J profile of PCn6:Kþ(1:0)/Al obviously increased
in comparison with that of the device without this layer
(Figure 1A). When K2CO3 was blended into PCn6 with a
Cn6:Kþ mole ratio of 1:1 or 1:3, the corresponding device
exhibited remarkably a higher J profile than PCn6:Kþ(1:0)/Al.
For example, J of PCn6:Kþ(1:3)/Al at 6 V was larger than those of
PCn6:Kþ(1:1)/Al, PCn6:Kþ(1:0)/Al, and the device without an
EIꢀHB layer by factors of 1.2, 74, and 1700, respectively. Similarly,
the B profiles followed the same trend (Figure 1B), and PCn6:
Kþ(1:3)/Al gaveBmax = 26 500 cd/m2, which is dramatically higher
than that from the device without an EIꢀHB layer (12 cd/m2). In
addition, Von decreased dramatically, from 10 V (no EIꢀHB layer)
to 5.8 V [PCn6:Kþ(1:0)/Al] to 3.6 V [PCn6:Kþ(1:1 or 1:3)/Al].
Also, PCn6:Kþ(1:x)/Al exhibited a much higher ηL than the device
without this layer, reaching the highest value (2.03 cd/A) at the
Cn6:Kþ mole ratio of 1:3. Obviously, the device performance with
Al as the cathode was remarkably enhanced when the water/
methanol-soluble PCn6:Kþ was inserted as the EIL. Because no
change was made on the ITO/PEDOT anode, the enhanced device
performance can be attributed to the increased electron current
density resulting from the reduction of the electron-injection barrier
provided by PCn6:Kþ(1:x) layers through the formation of an
interfacial dipole (Text S1 and Figure S1).
Figure 1. Characteristic PLED curves of (A) current density and (B)
brightness vs voltage and (C) luminous efficiency vs current density for
β-PFO-based devices without and with EIꢀHB layers. The device
structure was ITO/PEDOT (25 nm)/β-PFO (120 nm)/[without/with
EIꢀHB layer (20 nm)]/Al (60 nm).
Since the open-circuit voltage (Voc) of a device reflects the
built-in potential (the difference in work function between anode
and cathode) across its anode/EML/cathode junction, it can be
used to investigate the change in the electron-injection barrier of
devices due to the formation of an interfacial dipole with or
without insertion of an additional layer between the EML and
cathode.2,12 In order to demonstrate that the chelated Kþ plays
another role in barrier reduction, we plotted the reduced current
density J/Jbare-Al (where the subscript “bare-Al” refers to a device
with only Al as the cathode and without an EIꢀHB layer) versus
Voc ꢀ Voc,bare-Al [Voc values for PCn6:Kþ(1:x)/Al and PCn5:
Kþ(1:x)/Al were taken from Figures S1 and S5, respectively]. As
shown in Figure 2, J/Jbare-Al dramatically increased from 23 with
only PCn6 to 1411 with PCn6:Kþ(1:1). However, the corre-
sponding Voc differences (relative to Voc,bare-Al) increased only
9635
dx.doi.org/10.1021/ja111521p |J. Am. Chem. Soc. 2011, 133, 9634–9637

Journal of the American Chemical Society
COMMUNICATION
from 0.83 to 1.01 V. Obviously, the added Kþ plays another role
in barrier reduction in addition to the formation of an interfacial
dipole because it causes an increase in J/Jbare-Al larger than that
expected from the increase in the Voc difference. The additional
function is attributed to the formation of a pseudometallic state
of Kþ when Kþ is chelated by Cn6 [as revealed by X-ray
photoelectron spectroscopy (XPS) data below] that provides
an intermediate step beneficial for electron injection from the
cathode to the EIL. The dramatic increase of J/Jbare-Al also
happens in the case of PCn5:Kþ(1:x)/Al but to a lesser extent,
indicating that the inference of a pseudometallic state that
enhances electron injection is reasonable. In addition, the
J/Jbare-Al of PCn6:Kþ(1:3)/Al (1700) is larger than those with-
out Kþ intercalation calculated from the data in refs 2, 3, 5, and 6,
indicating the extraordinary effect of the pseudometallic state on
the enhancement of electron injection.
Figure 2. Reduced current density (J/Jbare-Al) vs Voc ꢀ Voc,bare-Al for
PCn6:Kþ(1:x), PCn5:Kþ(1:x), and literature data. Values in parentheses
are the electric fields at which the J values were recorded.
For further effective utilization of injected holes from the anode,
PEO was incorporated into the EIL by blending to provide the HB
function (Text S2). At a PCn6:PEO mass ratio of 1:0.75, the device
performance of PCn6:Kþ(1:3)/Al with the color coordinates
CIEx,y (0.160, 0.113) was further enhanced to Bmax = 54800 cd/m2
and ηL = 6.14 cd/A (ηext = 5.42%), which are the highest values
reported to date for deep-blue PLEDs with a conjugated polymer as
the EML (Text S3 and Figure S2). Further increases in the PEO
content led to a decrease in device performance. For PCn6:PEO-
(1:2)/Al, Bmax and ηL decreased to 8900 cd/m2 and 1.65 cd/A,
respectively, as a result of the fact that too many holes were blocked
and fewer electrons were injected, as indicated by the lower current
density than for PCn6:Kþ(1:3)/Al (e.g., by a factor of 2.1 at 6 V;
Figure 1A). On the other hand, the EL spectra of the devices with
EIꢀHB layers and that of the device based on the PCn6:Kþ(1:3)
film as the EML were quite different, demonstrating that the
EIꢀHB layer does not emit light and that the undesired green
emission does not grow during operation for the devices with
EIꢀHB layers (Text S4 and Figure S3). In addition, chelation of
PCn6 and PCn5 to other metal ions with different sizes (Liþ,
Naþ, and Csþ) could also enhance device performance as in the
cases of chelation to Kþ (Text S5 and Tables S1 and S2).
The evidence for the formation of the pseudometallic state is
presented below. We performed XPS measurements on PCn6:
Kþ(1:1) and -(1:3) films to investigate the interaction between
Cn6 and Kþ in the Cn6/Kþ complex. As shown in Figure 3, the
characteristic electron binding energies of K 2p3/2 and K 2p1/2
in K2CO3 are 294.11 and 296.91 eV, respectively. In the Cn6/
Kþ complex, these two binding energies shifted by 0.5 eV to the
lower values 293.61 and 296.41 eV, respectively. This indicates
that Kþ receives extra electrons from the oxygen atoms in Cn6,
causing its electronic states (K 2p3/2 and K 2p1/2) to shift
Figure 3. XPS spectra (K 2p) of K2CO3, PCn6:Kþ(1:1), and PCn6:
Kþ(1:3) films.
Cn6:Kþ(1:3)/Al (2.36 V, Figure S1) than for Cn6:Kþ(1:1)/Al
(2.25 V, Figure S1).
However, the lowering of the electron-injection barrier by the
EILs cannot completely explain the difference in performance
between devices with Cn6/Kþ and Cn5/Kþ (Text S7 and
Figures S5 and S6). We found that facilitating electron transport
in these EILs is also important for increasing the electron current
density,15 and the stronger interaction between Kþ and Cn6 can
form more Kþ channels across the EIL for electron transport
(Text S7 and Figures S7 and S8). Figure 4 illustrates the K 2p3/2
binding energies of various K/Kþ states, and one can see that the
interaction between Cn6 and Kþ is stronger than the Cn5/Kþ
and PEO/Kþ interactions because its binding energy is closer to
that of potassium metal than the others (Text S8 and Figure S9).
Therefore, we propose the working mechanism of the EIꢀHB
layer as illustrated in Figure 5. The chelated Kþ can further
reduce the electron-injection barrier through the formation of an
interfacial dipole [e.g., the barrier was reduced by 1.12 eV for
insertion of the PCn6:Kþ(1:3) layer relative to the value without
the layer] and the establishment of an intermediate step for
electron injection. Therefore, electrons can be directly injected
into the backbone of PCn6 or chelated Kþ. Because the lowest
unoccupied molecular orbital (LUMO) of PCn6 should be closer
to the vacuum level than the energy level of chelated Kþ, one can
expect that the electrons originally injected into the backbone of
toward those of potassium metal (K 2p3/2, 292.60 eV; K 2p1/2
,
295.50 eV).13 For the case of the dibenzo-18-crown-6/Kþ
complex, the two K 2p binding energies were found to be
independent of the counterion (Clꢀ, Brꢀ, or Iꢀ), implying
that electron donation from the lone-pair electrons of the oxygen
atoms can effectively stabilize Kþ.14 This new state of Kþ is thus
termed a “pseudometallic state” and is expected to provide a
bridge for electron injection and transport between the high-
work-function Al metal and the EML. In addition, from these
XPS results, the fractions of Cn6 chelated to Kþ were determined
by deconvolution of the K 2p signal peaks to be 39.5 and
94.7% for the mole ratios 1:1 and 1:3, respectively (Text S6
and Figure S4). The latter is higher than the former by a
factor of 2.4, accounting for the larger observed Voc value for
9636
dx.doi.org/10.1021/ja111521p |J. Am. Chem. Soc. 2011, 133, 9634–9637

Journal of the American Chemical Society
COMMUNICATION
for deep-blue-emission PLEDs, which allows the use of environ-
mentally stable Al as the cathode because electron donation from
the crown ether (Cn6) can reduce Kþ to a stable “pseudometallic
state”. Furthermore, when PEO was blended into the EIL (at a
Cn6:Kþ mole ratio of 1:3) to provide a hole-blocking effect, the
device exhibited the highest performance reported to date for
deep-blue emission, Bmax = 54 800 cd/m2 and ηext (ηL) = 5.42%
(6.14 cd/A), which are much higher even than those obtained
using CsF/Al as the cathode (34 326 cd/m2, 3.33%, and 3.85 cd/
A). The introduction of a stable pseudometallic state as an
intermediate step for effective electron injection and transport
opens a broad avenue leading toward industrialization of PLEDs.
’ ASSOCIATED CONTENT
Figure 4. K 2p3/2 binding energies (in eV) of various K/Kþ states.
S
Supporting Information. Description of experimental
b
data, syntheses of monomers and polymers, chelation, device
fabrication and instrumentation. This material is available free of
charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
sachen@che.nthu.edu.tw
’ ACKNOWLEDGMENT
The authors gratefully acknowledge the financial support from
National Science Council (Project 96-2752-E-007-008-PAE) and
kind help by Mr. Kun-Wei Tsai on the PLQE of the PFO film.
’ REFERENCES
(1) Huang, F.; Wu, H. B.; Wang, D. L.; Yang, W.; Cao, Y. Chem. Mater.
2004, 16, 708.
(2) Wu, H.; Huang, F.; Mo, Y.; Yang, W.; Wang, D.; Peng, J.; Cao, Y.
Adv. Mater. 2004, 16, 1826.
Figure 5. Schematic illustration of the working mechanism proposed
in this study. The numbers are energy levels in eV taken from the
literature.11,16 Δ denotes the shift of the vacuum level.
(3) Huang, F.; Niu, Y. H.; Zhang, Y.; Ka, J. W.; Liu, M. S.; Jen, A. K. Y.
Adv. Mater. 2007, 19, 2010.
PCn6 will spontaneously hop to the chelated Kþ. The chelated Kþ
can form electron-transport channels through which the electrons
can migrate to the EILꢀEML interface. The electrons can then
easily hop to the EML because the chelated Kþ is in the pseudo-
metallic state resembling potassium metal. In short, the existence
of chelated Kþ can facilitate electron injection from the Al cathode
(work function = 4.28 eV)16 to the EML (its LUMO level is
2.86 eV).11 In addition, PEO provides HB functionality to increase
the holeꢀelectron recombination fraction in the EML and there-
foreenhance thedevice performance. Thisproposedmechanism is
further supported by the fact that including the EIꢀHB layer was
useful when Au was used as the cathode (Text S9 and Table S3).
Notably, this mechanism for the enhancement of device perfor-
mance is different from that in polymer light-emitting electrochemi-
cal cells (PLECs), in which the generation of n- and p-doped
conjugated polymer layers at two electrode interfaces by addition of
an ionic salt under an electric field is involved.17,18 In the present case,
the ηL values for devices with Au as the cathode were lower than
those with Al as the cathode by a factor of 30ꢀ70(TextS9andTable
S3), indicating that the device performance based on the EIꢀHB
layers strongly depends on the cathode material used. For LECs, the
device performance was independent of the electrode material used.
In summary, we have presented the use of water/methanol-
soluble polyfluorene grafted with Cn6 chelated to Kþ as an EIL
(4) Huang, F.; Zhang, Y.; Liu, M. S.; Jen, A. K. Y. Adv. Funct. Mater.
2009, 19, 2457.
(5) Oh, S. H.; Vak, D.; Na, S. I.; Lee, T. W.; Kim, D. Y. Adv. Mater.
2008, 20, 1624.
(6) Yang, R.; Wu, H.; Cao, Y.; Bazan, G. C. J. Am. Chem. Soc. 2006,
128, 14422.
(7) Hoven, C. V.; Garcia, A.; Bazan, G. C.; Nguyen, T. Q. Adv. Mater.
2008, 20, 1.
(8) Yu, J. M.;Chen, Y.J. Polym. Sci., Part A: Polym. Chem. 2009, 47, 2985.
(9) Pedersen, C. J. J. Am. Chem. Soc. 1967, 89, 2495.
(10) Pedersen, C. J. Angew. Chem., Int. Ed. Engl. 1988, 27, 1021.
(11) Lu, H. H.; Liu, C. Y.; Chang, C. H.; Chen, S. A. Adv. Mater. 2007,
19, 2574.
(12) Yu, L. S.; Chen, S. A. Adv. Mater. 2004, 16, 744.
(13) Schulze, M.; Reissner, R.; Lorenz, M. Fresenius' J. Anal. Chem.
1999, 365, 178.
ꢁ
(14) Bohman, O.; Ahlberg, P.; Martensson, N.; Siegbahn, K. Phys. Scr.
1977, 16, 355.
(15) Huang, F.; Shih, P. I.; Shu, C. F.; Chi, Y.; Jen, A. K. Y. Adv. Mater.
2009, 21, 361.
(16) Chen, S. A.;Lu, H. H.;Huang, C. W. Adv. Polym. Sci. 2008, 212, 49.
(17) Pei, Q.; Yu, G.; Zhang, C.; Yang, Y.; Heeger, A. J. Science 1995,
269, 1086.
(18) Pei, Q.; Yang, Y.; Yu, G.; Zhang, C.; Heeger, A. J. J. Am. Chem.
Soc. 1996, 118, 3922.
9637
dx.doi.org/10.1021/ja111521p |J. Am. Chem. Soc. 2011, 133, 9634–9637