A comparison of CuO and Cu 2 O hole-injection layers for low voltage organic

devices
G. B. Murdoch, M. Greiner, M. G. Helander, Z. B. Wang, and Z. H. Lu

Citation: Applied Physics Letters 93, 083309 (2008); doi: 10.1063/1.2966140

View online: http://dx.doi.org/10.1063/1.2966140
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/93/8?ver=pdfcov
Published by the AIP Publishing

Articles you may be interested in

Pressure-assisted fabrication of organic light emitting diodes with MoO3 hole-injection layer materials
J. Appl. Phys. 115, 233703 (2014); 10.1063/1.4881780

Improved hole injection and transport of organic light-emitting devices with an efficient p-doped hole-injection
layer
Appl. Phys. Lett. 95, 263303 (2009); 10.1063/1.3279142

Current efficiency in organic light-emitting diodes with a hole-injection layer

Appl. Phys. Lett. 93, 093306 (2008); 10.1063/1.2978349

Polyfluorene-based light-emitting diodes with an azide photocross-linked poly(3,4-ethylene

dioxythiophene):(polystyrene sulfonic acid) hole-injecting layer
Appl. Phys. Lett. 92, 103308 (2008); 10.1063/1.2892685

Vanadium-doped indium tin oxide as hole-injection layer in organic light-emitting devices

Appl. Phys. Lett. 87, 243510 (2005); 10.1063/1.2137892

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:
209.172.228.248 On: Sun, 14 Dec 2014 00:13:28
 APPLIED PHYSICS LETTERS 93, 083309 共2008兲

A comparison of CuO and Cu2O hole-injection layers for low voltage

organic devices
G. B. Murdoch,a兲 M. Greiner, M. G. Helander, Z. B. Wang, and Z. H. Lub兲
Department of Materials Science and Engineering, University of Toronto, 184 College St., Toronto,
Ontario M5S 3E4, Canada
共Received 30 April 2008; accepted 6 July 2008; published online 29 August 2008兲
Cu2O and CuO have been grown with an aim to reduce junction electrical resistance when
interfaced with N , N⬘-bis共1-naphthyl兲-N , N⬘-diphenyl-1 , 1⬘ biphenyl 4 , 4⬘-diamine 共NPB兲. Organic
light-emitting diodes employing Cu/CuO anodes have equivalent driving voltages as devices made
with indium tin oxide. Hole-injection barriers are calculated from current-voltage characteristics of
CuO/NPB/Cu and Cu2O / NPB/ Cu devices via theoretical simulation. Photoelectron spectroscopies
are used to measure oxide valence band spectra, interfacial dipole formation, and band bending
during in situ sequential deposition of NPB on each oxide. Calculated hole-injection barriers and
those derived from photoemission results accord well, explaining the superior hole injection at the
CuO-NPB interface. © 2008 American Institute of Physics. 关DOI: 10.1063/1.2966140兴

With steady adoption for niche markets such as flat-
panel displays, organic electronics are now actively re-
searched for other applications such as organic solid-state
lighting, thin film transistors, memory units, and photovolta-
ics. Indium tin oxide 共ITO兲 has been used ubiquitously as an
anode material despite its ever-rising cost and technical chal-
lenges in current-driven devices such as organic light-
emitting diodes 共OLEDs兲. Recently, the use of metal oxides
as hole-injection layers 共HILs兲 for anodes has attracted con-
siderable interest.1–4 In particular, the Cu/ CuxO system
has shown promise as an efficient anode contact for
OLEDs,5,6 organic thin film transistors,7 and photovoltaics.8
In this work, we report that OLEDs based
on N , N⬘-bis共1-naphthyl兲-N , N⬘-diphenyl-1 , 1⬘ biphenyl
4 , 4⬘-diamine 共NPB兲 employing ultraviolet 共UV兲-ozone pro-
duced CuO, as opposed to Cu2O as HIL, exhibit significantly
better device performance with particularly low-driving volt-
age. By using UV photoelectron spectroscopy 共UPS兲, x-ray
photoelectron spectroscopy 共XPS兲, and in situ deposition of
NPB, we obtain oxide band structure and interface dipole
energy for both oxides. Hole-injection barriers into NPB de-
rived from UPS results agree well with transport modeling of
single carrier devices, explaining device performance differ-
ences in terms of the hole-injection ability of each material.
OLED devices were fabricated on glass substrates
cleaned in acetone, methanol, and de-ionized water, followed
by 15 min UV-ozone treatment. Reference devices were fab-
ricated on ITO coated glass 共⬃15 ⍀ / 䊐兲 cleaned in the same
manner. Hole-only devices were fabricated on Si共100兲 with
2000 nm furnace oxides for electrical isolation. Metal and
organic films were thermally evaporated in separate cham-
bers with base pressures of ⬃10−7 and ⬃10−8 Torr, respec-
tively. Film thicknesses were monitored using calibrated
quartz crystal monitors. As will be shown, CuO and Cu2O
oxide layers were produced by ex situ oxidation of freshly
deposited Cu films for 60 min by UV-ozone treatment and
annealing in air at 100 ° C, respectively.
Immediately after oxidation, organic and cathode layers
were deposited to obtain the following OLED structure:
anode/NPB 共60 nm兲/tris-共8-hydroxyquinoline兲 aluminum
共Alq3兲 共45 nm兲/LiF 共1.5 nm兲/Al 共100 nm兲. The structure of
the hole-only devices was as follows: anode/NPB 共500
nm兲/Cu 共20 nm兲. The active area for all devices was 1
⫻ 2 mm2. Luminance and current density 共L-J-V兲 character-
istics were measured in ambient atmosphere using an
HP4140B picoampere meter and a Minolta LS-110 lumi-
nance meter. The perpendicular anode and cathode lines
were contacted approximately 2 cm away from the device
active area. XPS and UPS measurements were performed
using a PHI 5500 system with monochromated Al K␣ x-rays
共1486.6 eV兲 and He I UV 共21.2 eV兲. All spectra were cali-
brated by referencing to the Fermi edge of sputtered Au.
Figure 1 compares current density and luminance for
OLEDs made with semitransparent Cu/CuO and ITO anodes.
Cu film thickness prior to oxidation was optimized at 20 nm,
compromising between optical transparency and series sheet
resistance. The Cu/CuO anode device exhibited approxi-
mately equal driving voltage to the ITO baseline device de-
spite the lower transparency and larger sheet resistance of the
Cu/CuO anode 共25 ⍀ / 䊐 as opposed to 15 ⍀ / 䊐 for ITO兲.
The driving voltage at 8000 cd/ m2 luminance was 7.4 V for
the Cu/CuO device compared with 7.6 V for the ITO refer-
ence device. Past 7 V, the Cu/CuO device overtakes the ITO
device in both current density and luminance.

a兲
Electronic mail: graham.murdoch@utoronto.ca. FIG. 1. 共Color online兲 Current density vs voltage and luminance vs voltage
b兲
Electronic mail: zhenghong.lu@utoronto.ca. for OLEDs made with UV-oxidized Cu and ITO anodes.

0003-6951/2008/93共8兲/083309/3/$23.00
This article 93,is083309-1
is copyrighted as indicated in the article. Reuse of AIP content © 2008 American InstituteDownloaded
subject to the terms at: http://scitation.aip.org/termsconditions. of Physics to IP:
209.172.228.248 On: Sun, 14 Dec 2014 00:13:28
 083309-2 Murdoch et al.
FIG. 2. 共a兲 Cu 2p3/2 XPS and 共b兲 Cu LMM Auger spectra 共h␯
= 1486.6 eV兲 for 5 nm Cu films heated at 100 ° C for 60 min and UV ozone
treated for 60 min.
Previous reports of Cu oxide anodes6 or anode modifi-
cation layers5 did not distinguish any preferred Cu oxidation
state. This distinction can be made, however, by examining
both 2p3/2 XPS and LMM Auger spectra.9 XPS data are
shown in Fig. 2共a兲. The UV-ozone sample exhibits a satellite
and a main peak, characteristic of CuO. The heated sample
has no satellite and main peaks, consistent with either metal-
lic Cu or Cu2O. The Auger LMM peak position 关Fig. 2共b兲兴,
however, confirms the presence of Cu2O only.
The thicknesses of UV-ozone oxide was found to plateau
at ⬃6 nm after 60 min of UV-ozone treatment, as deter-
mined by XPS sputter profiling and stylus profilometry of the
sputter crater.
In order to elucidate the hole injecting properties of CuO
and Cu2O, single carrier devices were fabricated. Figure 3
shows the room temperature J-V characteristics of hole-only
devices made with Cu/CuO and Cu/ Cu2O anode structures.
Qualitatively, it is seen that the Cu/CuO anode exhibits a
much larger hole current. A useful model approximating
charge injection into organic materials was described by
Scott and Malliaras,10,11 and it has been used to fit the single
carrier device data. In this model, charge recombination at
the organic interface due to carrier hopping within an image
potential is used in combination with the principle of detailed
balance to calculate the electric field dependence of thermi-
FIG. 3. 共Color online兲 共a兲 Current density vs voltage for hole-only structures charging became significant after 25.6 nm, as seen by large
as follows: anode/NPB 共500 nm兲/Cu. The anodes used were Cu/CuO and
Cu/ Cu2O. Fits to the Scott and Malliaras 共see Refs. 10 and 11兲 injection
model are included as solid lines along with ␾b values. 共b兲 Current density
vs voltage and luminance vs voltage 共inset兲 for OLEDs made with CuO and
Cu2Oiscapped
This article ITO anodes.
copyrighted
as indicated in the article. Reuse of AIP content is subject to thefits of at:
209.172.228.248 On: Sun, 14 Dec 2014 00:13:28
Appl. Phys. Lett. 93, 083309 共2008兲
FIG. 4. 共Color online兲 He I UPS valence band spectra for 共a兲 CuO and 共b兲
Cu2O substrates with sequential NPB deposition up to a thickness of 12.8
nm. Work function shifts and C 1s core level shifts vs NPB thickness are
plotted for 共c兲 CuO substrate and 共d兲 Cu2O substrate.
onic injection. The field and temperature dependence of the
injection current is given by
J = 4e␺2N0E␮ exp共− ␾b/kbT兲f 1/2 , 共1兲
where f is the reduced electric field,
f = e3E/4␲␧k2bT2 , 共2兲
and
␺共f兲 = f −1 + f −1/2 − f −1共1 + 2f 1/2兲1/2 . 共3兲
In the above equations, E is the applied electric field, ␮ is the
field dependent carrier mobility, N0 is the density of charge-
able states in the organic film, ␾b is the barrier height, kB is
the Boltzmann constant, T is the Kelvin temperature, and ␧ is
the dielectric constant of the organic material.
The field dependent mobility and relative permittivity
for NPB deposited in our vacuum system were obtained
through time-of-flight and capacitance-voltage measure-
ments reported elsewhere.12 The density of chargeable states
in the highest occupied molecular orbital 共HOMO兲 共N0
= 1021 cm−3兲 is taken by assuming one state per molecule.
Figure 3共a兲 shows fits to the single carrier device data. Bar-
rier heights ␾b = 0.3 eV for the CuO/NPB interface and ␾b
= 0.8 eV for the Cu2O / NPB interface were extracted. Figure
3共b兲 shows J-V and L-V characteristics for OLEDs made
using ITO modified with 5 nm of either oxide of Cu. As
expected from the single carrier devices, capping with Cu2O
resulted in greatly reduced current density and luminance
when compared with CuO; this can be attributed to the re-
duction in hole-injection barrier height.
XPS and UPS were used to investigate the electronic
structure of the oxide/NPB interface for both types of oxides.
This combination of techniques allows one to distinguish
between interfacial dipole formation and band bending, in
addition to possible charging.13,14 Films of NPB were depos-
ited on the oxides by step-by-step deposition while obtaining
XPS and UPS spectra in situ after each deposition step. The
final NPB thickness used for analysis was 12.8 nm as sample
XPS peak shifts when switching the UPS photon flux on and
off. Figures 4共a兲 and 4共b兲 show He I valence band spectra for
the NPB deposition sequence on CuO and Cu2O substrates.
Linear
terms thehttp://scitation.aip.org/termsconditions.
substrate valence band features and back- to IP:
Downloaded
 083309-3 Murdoch et al. Appl. Phys. Lett. 93, 083309 共2008兲

grounds are used to determine the valence band maximum
共VBM兲 positions relative to the Fermi level 共EF兲. These rela-
tive positions, denoted ⌬Ev, are 0.1 eV in the case of CuO
and 0.3 eV in the case of Cu2O. With increasing NPB thick-
ness, NPB’s HOMO spectral feature becomes apparent. The
relevant hole-injection barrier is the energy difference be-
tween the oxide VBM and the HOMO maximum 共HM兲 be-
fore band bending. The HM is taken as the low binding
energy cutoff of the HOMO feature.13–15 Referring to the
inset of Fig. 3共a兲, we have
␾b = EHM − Vb − ⌬Ev , 共4兲
where EHM is the position of HM relative to EF and Vb is the
degree of band bending. At 12.8 nm NPB thickness, EHM is
0.9 eV for the CuO system and 1.1 eV for the Cu2O system.
In order to determine the amount of band bending in the
NPB, corresponding shifts in the C 1s XPS peaks were
monitored for both oxide systems. In addition, work function
shifts were measured via UPS secondary cutoff. Figures 4共c兲
and 4共d兲 show that at an NPB thickness of 12.8 nm the
amount of band bending was 0.4 and 0.1 eV for the CuO and
Cu2O systems, respectively. The work function of the CuO
substrate was measured to be 5.3 eV and that of the Cu2O
substrate was 4.6 eV, which agree with literature values.16,17
Additionally, deviation of the C 1s shift from the UPS work
function shift at low NPB coverage for both material systems
indicates the presence of an interface dipole ⌬.13 The values
of ⌬ are found to be 0.4 and 0.2 eV for the CuO and Cu2O
systems, respectively. Using Eq. 共4兲, hole-injection barriers
are calculated to be 0.4 eV for the CuO/NPB interface and
0.7 eV for the Cu2O / NPB interface, in good agreement with
the previous transport calculations.
In summary, it is found that UV-ozone treated Cu acts as
a suitable anode for OLEDs, imparting low-driving voltage
and high luminous output despite its moderate transparency.
Low cost, high conductivity, easy patternability, and low
thermal budget make such ozone treated Cu-based anodes
attractive for use in commercial applications. Moreover, we
have identified CuO as the oxidation state of Cu most suited
for hole injection into NPB. Transport modeling of single
carrier devices yields calculated hole-injection barrier
heights that harmonize well with the observed electronic
structure investigated via photoemission, showing a signifi-
cantly smaller barrier at the CuO/NPB interface. Although
this study has focused exclusively on hole-injection into
NPB, the Cu/CuO system is a potential candidate for low
junction resistance anode contacts for other organic materi-
als.
1
I. M. Chan and F. C. Hong, Thin Solid Films 450, 304 共2004兲.
2
S. Y. Kim, J. M. Baik, H. K. Yu, and J. L. Lee, J. Appl. Phys. 98, 093707
共2005兲.
3
J. Li, M. Yahiro, K. Ishida, H. Yamada, and K. Matsushige, Synth. Met.
151, 141 共2005兲.
4
C. F. Qiu, Z. L. Xie, H. Y. Chen, M. Wong, and H. S. Kwok, J. Appl. Phys.
93, 3253 共2003兲.
5
W. Hu, K. Manabe, T. Furukawa, and M. Matsumura, Appl. Phys. Lett.
80, 2640 共2002兲.
6
S. Wang, T. Osasa, and M. Matsumura, Jpn. J. Appl. Phys., Part 1 45,
8894 共2006兲.
7
C. Di, G. Yu, Y. Liu, Y. Guo, Y. Wang, W. Wu, and D. Zhu, Adv. Mater.
共Weinheim, Ger.兲 20, 1286 共2008兲.
8
A. Mittiga, E. Salza, F. Sarto, M. Tucci, and R. Vasanthi, Appl. Phys. Lett.
88, 163502 共2006兲.
9
J. Ghijsen, L. H. Tjeng, J. van Elp, H. Eskes, J. Westerink, G. A. Sa-
watzky, and M. T. Czyzyk, Phys. Rev. B 38, 11322 共1988兲.
10
J. C. Scott and G. G. Malliaras, Chem. Phys. Lett. 299, 115 共1999兲.
11
J. C. Scott, J. Vac. Sci. Technol. A 21, 521 共2003兲.
12
S. W. Tsang, Z. H. Lu, and Y. Tao, Appl. Phys. Lett. 90, 132115 共2007兲.
13
R. Schlaf, C. D. Merritt, L. A. Crisafulli, and Z. H. Kafafi, J. Appl. Phys.
86, 5678 共1999兲.
14
R. Schlaf, P. G. Schroeder, M. W. Nelson, and B. A. Parkinson, J. Appl.
Phys. 86, 1499 共1999兲.
15
S. Y. Kim and J. L. Lee, Appl. Phys. Lett. 89, 223515 共2006兲.
16
F. P. Koffyberg and F. A. Benko, J. Appl. Phys. 53, 1173 共1982兲.
17
S. S. Jeong, A. Mittiga, E. Salza, A. Masci, and S. Passerini, Electrochim.
Acta 53, 2226 共2008兲.

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:
209.172.228.248 On: Sun, 14 Dec 2014 00:13:28