Direct evidence of n-type doping in organic light-emitting devices: N free Cs doping from CsN3

Article abstract of DOI:10.1063/1.4718017

Cesium azide (CsN₃) is confirmed to be decomposed during thermal evaporation. Only Cs could be deposited on tris(8-hydroxyquinolinato)aluminum (Alq₃) and n-type doping is easily achieved. Organic light-emitting devices with CsN₃ show highly improved current density-luminance- voltage characteristics compared to the control device without CsN₃. To understand the origin of the improvements, in situ x-ray and UV photoemission spectroscopy measurements were carried out and a remarkable reduction in electron injection barrier is verified with successive deposition of Al on CsN₃ on Alq₃. CsN₃ has a potential as alternative to doping the electron transport layer by replacing the direct deposition of alkali metals.

Full text of DOI:10.1063/1.4718017

Direct evidence of n-type doping in organic light-emitting devices: N free Cs doping
from CsN3
Jeihyun Lee, Hyunbok Lee, Pyungeun Jeon, Kwangho Jeong, Tae Gun Kim, Jeong Won Kim, and Yeonjin Yi
Citation: Applied Physics Letters 100, 203301 (2012); doi: 10.1063/1.4718017
View online: http://dx.doi.org/10.1063/1.4718017
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/100/20?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Energy transfer between host and dopant molecules in blue organic light-emitting devices
J. Appl. Phys. 110, 034501 (2011); 10.1063/1.3614493
Suppression of efficiency roll off in blue phosphorescent organic light-emitting devices using double emission
layers with additional carrier-transporting material
Appl. Phys. Lett. 94, 083506 (2009); 10.1063/1.3083558
Electronic structures and electron-injection mechanisms of cesium-carbonate-incorporated cathode structures for
organic light-emitting devices
Appl. Phys. Lett. 88, 152104 (2006); 10.1063/1.2192982
Mixing effect of hole-injecting and hole-transporting materials on the performance and lifetime of organic light-
emitting devices
Appl. Phys. Lett. 88, 043504 (2006); 10.1063/1.2168014
Photoemission study of hole-injection enhancement in organic electroluminescent devices with Au/CF x anode
Appl. Phys. Lett. 84, 73 (2004); 10.1063/1.1637945
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.21.35.191 On: Thu, 18 Dec 2014 21:59:21

APPLIED PHYSICS LETTERS 100, 203301 (2012)
Direct evidence of n-type doping in organic light-emitting devices: N free Cs
doping from CsN3
Jeihyun Lee,1 Hyunbok Lee,1 Pyungeun Jeon,1 Kwangho Jeong,1 Tae Gun Kim,2
Jeong Won Kim,3 and Yeonjin Yi1,a)
1Institute of Physics and Applied Physics, Yonsei University, 50 Yonsei-ro, Seodaemoon-gu, Seoul 120-749,
South Korea
2University of Science and Technology, 217 Gajeong-ro, Yuseong-gu, Daejeon 305-350, South Korea
3Korea Research Institute of Standards and Science, 267 Gajeong-ro, Yuseong-gu, Daejeon 305-340,
South Korea
(Received 13 January 2012; accepted 28 April 2012; published online 14 May 2012)
Cesium azide (CsN3) is confirmed to be decomposed during thermal evaporation. Only Cs could
be deposited on tris(8-hydroxyquinolinato)aluminum (Alq3) and n-type doping is easily
achieved. Organic light-emitting devices with CsN3 show highly improved current
density-luminance-voltage characteristics compared to the control device without CsN3. To
understand the origin of the improvements, in situ x-ray and UV photoemission spectroscopy
measurements were carried out and a remarkable reduction in electron injection barrier is verified
with successive deposition of Al on CsN3 on Alq3. CsN3 has a potential as alternative to doping the
C
V
electron transport layer by replacing the direct deposition of alkali metals.
Institute of Physics. [http://dx.doi.org/10.1063/1.4718017]
2012 American
Organic materials have expanded their application in
areas such as organic light-emitting devices (OLEDs), or-
ganic thin-film transistors, sensors, and organic solar cells.
OLEDs are the most widely studied devices since they are al-
ready matured as a commercial product. Numerous studies
have been conducted in many aspects including materials,
carrier injection and transport, luminance and luminous effi-
ciencies. It has been a basic strategy to fabricate the OLEDs
with multilayer structures after Tang et al. showed efficient
devices with multilayered architecture.1 Such successes of
multilayers used to be achieved with the help of adequate
energy level alignments at interfaces.2,3 Therefore, interfa-
cial energy level alignments in devices have been a critical
issue to achieve high performance of OLEDs because charge
carriers have to overcome the energy barriers at multilayer
interfaces from respective electrodes to the emissive layer.
Mostly, electron injection has been an issue because the typi-
cal electron mobility and injection barrier are unfavorable
compared to those of holes. To reduce this mismatch
between two carriers, electron injection barriers should be
reduced at the interface and thus this is one of the most
widely studied issues.
mal evaporator at moderate temperature, it would ease the
difficulties of n-type doping with alkali metals.
Cesium azide (CsN3) is also an alkali metal compound
but it has been thought to possess the unique property that it
decomposes into Cs and N2 during the thermal evaporation
process, and so it was applied to dope an ETL.11 Very
recently, Yang et al. used it as an n-type dopant in an inter-
connector layer.12 Another advantage of CsN3 is the low
evaporation temperature (ꢀ300 ꢁC) which is similar to that
of organic materials. In addition, it can be deposited with a
conventional thermal evaporator and it is stable in ambient.
Therefore, only the Cs metal could be deposited from a con-
ventional thermal evaporator with ease of handling. These
unique properties make CsN3 a potential alternative as an al-
kali metal source for device fabrication.
Surprisingly, in spite of recent frequent applications of
CsN3, there are no direct/detailed reports that confirm the N
free Cs deposition and its n-type doping mechanism except
simplistic measurements.11 It should be confirmed prior to
the device application whether CsN3 is actually decomposed
and only Cs is deposited on the substrate and improves de-
vice characteristics as well. To address such issues practi-
cally, we fabricated OLEDs with CsN3 as an n-type dopant
and measured photoemission spectra at the interface of ETL/
CsN3/cathode to see the interfacial electronic structures as
well as the N free Cs deposition from a CsN3 source. CsN3
insertion showed highly improved device characteristics and
lowered the electron injection barrier just like the Cs depos-
ited directly from an alkali metal dispenser.
There are two conventional methods to reduce electron
injection barriers, which are the insertion of either an alkali
metal itself4–7 or an alkali metal compound8–10 between the
cathode and the electron transport layer (ETL). These meth-
ods have several weak points: (1) difficulties in deposition
and handling due to high reactivity, (2) limited deposition
rate, e.g., the alkali metal dispenser, (3) insulating nature of
the compound, (4) high evaporation temperature which
would not be compatible with pre-deposited organic layers.
If alkali metals could be deposited from a conventional ther-
OLEDs with CsN3 were fabricated with conventional
structures of indium-tin-oxide (ITO, 150 nm)/N,N0-diphenyl-
N,N0-bis(1-naphthyl)(1,10-biphenyl)-4,40diamine (NPB, 60 nm)
/tris(8-hydroxyquinolinato)aluminum (Alq3, 84nm)/CsN3 (0,
0.3, 0.5, 0.7, 1.0, 1.3 nm)/Al (100 nm). ITO patterned glass
substrates were cleaned successively with ultrasonication in de-
tergent, methanol, acetone, and deionized water. Organic
a)Author to whom correspondence should be addressed. Electronic mail:
yeonjin@yonsei.ac.kr.
C
V
0003-6951/2012/100(20)/203301/4/$30.00
100, 203301-1
2012 American Institute of Physics
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.21.35.191 On: Thu, 18 Dec 2014 21:59:21

203301-2
Lee et al.
Appl. Phys. Lett. 100, 203301 (2012)
FIG. 1. (a) XPS spectra of the N 1s level. There are no
traces of nitrogen before or after the CsN3 deposition. (b)
XPS spectra of the Cs 3d level measured after CsN3
(0.7 nm) deposition. It shows the metallic Cs deposited
on the substrate.
materials and CsN3 (Aldrich, purity > 99.99%) were thermally
evaporated with a pressure of 1.0 ꢂ 10ꢃ8 and 1.0ꢂ 10ꢃ5 Torr
at the rates of 0.1 and 0.02 nm/s, respectively. The Al cathode
was finally deposited at the rate of > 0.1 nm/s with a pressure
of 1.0ꢂ 10ꢃ6 Torr. Current density-voltage (J-V) and
luminance-voltage (L-V) were measured with a Keithley 237
source measure unit and an optimized Si photodiode with
Keithley 2400 source meter. All measurements were done in
dry nitrogen ambiance at room temperature.
X-ray and UV photoemission spectroscopy (XPS and
UPS) measurements were performed in situ to understand
the working mechanism of CsN3. A pristine Alq3 film was
prepared on a clean ITO substrate and CsN3 and Al were de-
posited on Alq3 in a stepwise manner. Before and after each
deposition step, the sample was transferred to the analysis
chamber without breaking the vacuum and XPS and UPS
measurements were conducted. The base pressure of the dep-
osition and analysis chamber were 2 ꢂ 10ꢃ8 and
5 ꢂ 10ꢃ10 Torr, respectively. A PSP RESOLVE 120 electron
spectrometer was used with UV (He I, 21.22 eV) and stand-
ard x-ray (Mg Ka, 1253.6 eV) sources. During UPS measure-
ment, a ꢃ10 V bias was applied to the sample to obtain the
true sample work function.
First, we checked if the CsN3 is decomposed and only Cs
is deposited. CsN3 was deposited on an ITO substrate and in
situ XPS measurement confirmed the decomposition. Fig. 1(a)
shows the N 1s core level spectra before and after the CsN3
deposition on ITO. No trace of N-related emission is observed
upon the CsN3 deposition. Instead Fig. 1(b) shows the clear
Cs 3d core level spectrum after the CsN3 deposition on ITO.
The 3d5/2 peak is located at 725.74 eV, which is close to the
value for metallic Cs, not CsN3.13 Therefore, it is demon-
strated clearly that the CsN3 decomposition occurs, nitrogen is
removed, and only metallic Cs is deposited on the substrate.
To see the effect of CsN3 as an n-type dopant in OLEDs,
J-V and L-V measurements were carried out. Figs. 2(a) and
2(b) show the measured J-V and L-V characteristics, respec-
tively. We varied the CsN3 thickness to find optimal device
performance and a CsN3 layer of 0.7 nm showed the best
characteristics in both current density and luminance. At
11.95 V, the optimal device shows current density and lumi-
nance of 2495.6 mA/cm2 and 3387.3 cd/m2, respectively.
FIG. 2. (a) Current density versus voltage (b) luminance versus voltage (c)
luminous efficiency versus current density. All data were obtained from the
device of ITO (150 nm)/NPB (60 nm)/Alq3 (84 nm)/CsN3 (x nm)/Al
(100 nm) with different CsN3 thickness.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.21.35.191 On: Thu, 18 Dec 2014 21:59:21

203301-3
Lee et al.
Appl. Phys. Lett. 100, 203301 (2012)
FIG. 3. UPS spectra of (a) normalized
secondary electron cutoff and (b)
HOMO region after background re-
moval during the interface formation of
Alq3/CsN3/Al (unit: eV).
Comparing them with those of a control device without
CsN3, CsN3 shows more than 20 and 70 times enhancements
in current density and luminance, respectively. This
enhancement can be explained by the typical n-type doping
with Cs deposition on Alq3 because only Cs is deposited on
Alq3 due to the CsN3 decomposition confirmed above. Fig.
2(c) shows luminous efficiency versus current density of the
same devices. The device with a CsN3 layer of 0.5 nm
showed better efficiency than that with a CsN3 layer of
0.7 nm because the charge imbalance between holes and
electrons is accomplished at the NPB/Alq3 interface. That is,
in the device with a CsN3 layer of 0.7 nm, the amount of
electrons far exceeds the amount of holes and the extra elec-
trons do not contribute to the recombination process. This
tendency of efficiency also confirms the efficient doping
with CsN3 clearly. The device performance would be much
improved if an appropriate hole injection layer is used to-
gether with the CsN3 dopant by improving charge balance.14
To understand the origin of the improvements in OLEDs
with CsN3, in situ UPS and XPS measurements were carried
out. We prepared a pristine Alq3 layer having sufficient
thickness (10 nm) to exclude any possible interacting region
with the ITO substrate. Then we studied the interface forma-
tion between Alq3 and Al upon the insertion of CsN3. The
multilayered structures of ITO/Alq3 (10.0 nm)/CsN3 (0.03,
0.1, 0.3, 0.7 nm)/Al (0.02, 0.07, 0.2, 0.7, 2.0 nm) were used
for the photoemission measurement. We chose a final CsN3
thickness of 0.7 nm as the OLED with the same thickness
shows the highest electron injection. We also measured the
control interface without CsN3 insertion (spectra not shown).
Fig. 3(a) shows the secondary electron cutoff (SEC)
region of the UPS spectra. The SEC shifts to higher binding
energies by 1.18 eV during the CsN3 deposition. It gradually
approaches to the work function of Al with an Al deposition,
resulting in total shift of 1.44 eV back to lower binding ener-
gies. Fig. 3(b) shows the highest occupied molecular orbital
(HOMO) region during the step-by-step deposition. The
HOMO onset of pristine Alq3 is seen at 2.03 eV below the
Fermi level. As soon as CsN3 is deposited on Alq3, all the
energy levels including the HOMO shift toward high binding
energy from the very first CsN3 deposition. The total HOMO
shift during the deposition of a CsN3 layer of 0.7 nm is
1.18 eV. This shift is the same as the total shift observed in
the SEC, implying a rigid level shift without an interface
dipole (eD). This rigid high-binding shift is contributed to
the n-type doping by charge donation of the deposited Cs.
Successive Al deposition shifts the HOMO level further due
to the extra charge transfer with Al. Al could interact with
Alq3 further through the thin Cs layer since a Cs layer of
0.7 nm is smaller than the one monolayer thickness of
Alq3.15 In addition to these shifts, a new gap state emerges in
the original gap of Alq3, indicating strong interactions with
the deposited Cs. The initial onset of the gap state is at
1.22 eV (determined from a CsN3 layer of 0.1 nm) below the
Fermi level, which shifts 0.40 and 0.37 eV toward high bind-
ing energies during CsN3 and Al deposition, respectively.
This n-type doping and gap state formation is consistent with
previous results with direct deposition of alkali metals.5,7
Therefore, CsN3 works in the same way as the Cs metal is
deposited directly. This confirms again the N free Cs deposi-
tion from the CsN3 source.
Combining all information obtained from spectral
changes, the energy level diagrams of Alq3/CsN3/Al and
Alq3/Al were drawn. In the diagram, the lowest unoccupied
molecular orbital (LUMO) level was estimated from the
transport gap of Alq3.16 Fig. 4(a) shows the control interface
of Alq3/Al (spectra not shown) and the LUMO level of Alq3
shifts downward by 0.89 eV with Al deposition. This LUMO
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.21.35.191 On: Thu, 18 Dec 2014 21:59:21

203301-4
Lee et al.
Appl. Phys. Lett. 100, 203301 (2012)
FIG. 4. Energy level diagrams of (a) Al
on Alq3 and (b) Al on CsN3 on Alq3
(unit: eV). EF is the Fermi level, Eion
the ionization energy, Evac the vacuum
level, eD the interface dipole, Vb the
level shift, Ue the electron injection bar-
rier, and UAl the work function of the
final surface.
shift (Vb) is not the same as the vacuum level shift, indicat-
ing the existence of an eD. These shifts result finally in an
electron injection barrier (Ue) of 0.71 eV. In Fig. 4(b), as
seen in Fig. 3, no eD is induced but a rigid level shift results
with CsN3 deposition. The level shift (Vb) caused by n-type
doping of Cs is 1.18 eV and the following Al deposition
makes a further downward level shift. As a result, the
final Ue is 0.15 eV, which is much smaller than that without
CsN3. Therefore, ohmic contact between the Al cathode and
Alq3 is achieved and the device characteristics are highly
improved.
and by a research project of the National Research Founda-
tion of Korea. (Grant No. 2011-0004748).
1C. W. Tang and S. A. VanSlyke, Appl. Phys. Lett. 51, 913 (1987).
2H. Ishii, K. Sugiyama, E. Ito, and K. Seki, Adv. Mater. 11, 605 (1999).
3A. Kahn, N. Koch, and W. Gao, J. Polym. Sci., Part B: Polym. Phys. 41,
2529 (2003).
4N. Johansson, T. Osada, S. Stafstro¨m, W. R. Salaneck, V. Parente, D. A.
dos Santos, X. Crispin, and J. L. Bre´das, J. Chem. Phys. 111, 2157 (1999).
5H. Ding and Y. Gao, Appl. Phys. Lett. 86, 213508 (2005).
6T. Oyamada, C. Maeda, H. Sasabe, and C. Adachi, Jpn. J. Appl. Phys.,
Part 2 42, L1535 (2003).
7M. K. Fung, S. L. Lai, S. W. Tong, S. N. Bao, C. S. Lee, and S. T. Lee,
Chem. Phys. Lett. 392, 40 (2004).
In summary, CsN3 was used as an n-type dopant in
OLEDs due to its unique properties. The decomposition of
CsN3 and N free Cs deposition on a substrate was confirmed.
OLEDs with CsN3 showed highly improved device charac-
teristics compared to the device without CsN3. This improve-
ment originates from the n-type doping effect of Cs
decomposed from CsN3. XPS and UPS measurements
clearly showed the level shift corresponding to the n-type
doping and this shift resulted in the reduction of the electron
injection barrier. Cs deposited from CsN3 works in an identi-
cal way as Cs directly deposited from a metallic source, e.g.,
alkali metal dispenser. These make CsN3 a potential source
for n-type doping in organic devices.
8L. Duan, Q. Liu, Y. Li, Y. Gao, G. Zhang, D. Zhang, L. Wang, and Y.
Qiu, J. Phys. Chem. C 113, 13386 (2009).
9M. G. Mason, C. W. Tang, L. S. Hung, P. Raychaudhuri, J. Madathil, D. J.
Giesen, L. Yan, Q. T. Le, Y. Gao, S. T. Lee, L. S. Liao, L. F. Cheng, W. R.
Salaneck, D. A. dos Santos, and J. L. Bre´das, J. Appl. Phys. 89, 2756 (2001).
10T. Wakimoto, Y. Fukuda, K. Nagayama, A. Yokoi, H. Nakada, and M.
Tsuchida, IEEE T. Electron Devices 44, 1245 (1997).
11K. S. Yook, S. O. Jeon, S. Y. Min, J. Y. Lee, H. J. Yang, T. Noh, S. K.
Kang, and T. W. Lee, Adv. Funct. Mater. 20, 1797 (2010).
12J. P. Yang, Y. Xiao, Y. H. Deng, S. Duhm, N. Ueno, S. T. Lee, Y. Q. Li,
and J. X. Tang, Adv. Funct. Mater. 22, 600 (2012).
13J. F. Moudler, W. F. Stickle, P. E. Sobol, and K. D. Bomben, Handbook of
X-Ray Photoemission Spectroscopy (Perkin-Elmer, Minnesota, 1992).
14H. You, Y. Dai, Z. Zhang, and D. Ma, J. Appl. Phys. 101, 026105 (2007).
15L. Yan, M. G. Mason, C. W. Tang, and Y. Gao, Appl. Surf. Sci. 175–176,
412 (2001).
16S. Krause, M. B. Casu, A. Scholl, and E. Umbach, New J. Phys. 10,
085001 (2008).
This work was supported by Brain Korea 21 (BK21)
project of the ministry of Education, Science and Technology
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.21.35.191 On: Thu, 18 Dec 2014 21:59:21