Improving operating lifetime of organic light-emitting diodes
with polycyclic aromatic hydrocarbons as aggregating
light-emitting-layer additives
Viktor V. Jarikov
a͒
Research & Development, Eastman Kodak Company, Rochester, New York 14650
͑Received 28 March 2006; accepted 12 May 2006; published online 5 July 2006͒
It is common in organic light-emitting diode technology to construct a light-emitting-layer ͑LEL͒
host with materials that resist luminescence-reducing aggregation, which is one of the common
reasons behind a phenomenon widely referred to as concentration quenching. However, if a host
material in its aggregated state has a substantial quantum yield of fluorescence ͑e.g., at least several
percents͒, it may yet be useful. We describe a group of aggregating flat and rigid polycyclic aromatic
hydrocarbons ͑PAHs͒ as LEL additives. These molecules readily form emissive aggregates when
added to the LEL. In the resulting devices, the aggregates show low-to-moderate external quantum
efficiencies ͑EQE͒ of 0.2%–1.3%. Significantly, the addition of these PAHs increases device
half-life ͑t
50
͒ 4–200 times, depending on the additive, up to 100 000 h upon operation at
40 mA/cm
2
. The lifetime increase occurs with many diverse classes of PAHs. The EQE can be
improved to 3.7% by further adding a proper dopant while maintaining the increased lifetime. A
possible link between the ability to aggregate and the lifetime increase is illustrated by comparing
aggregation-prone perylene and aggregation-resistant 2,5,8,11-tetra-t-butylperylene ͑TBP͒. Despite
the similarity between the two additives with respect to their initial device performance, perylene’s
stronger ability to aggregate correlates with the eight times longer half-life versus that for TBP.
© 2006 American Institute of Physics. ͓DOI: 10.1063/1.2214535͔
I. INTRODUCTION
Despite significant progress in efficiency, reliability, and
manufacturability of organic light-emitting diodes ͑OLEDs͒,
their operating lifetime remains a fundamental problem, a
high priority issue in research, and a factor that limits com-
mercial applications, such as flat panel displays. OLED deg-
radation may be divided into intrinsic, i.e., loss of electrolu-
minescent ͑EL͒ efficiency ͑“fade”͒ that is uniform across the
area and monotonic over hundreds or thousands of hours of
operation, and extrinsic, e.g., dark spot and short circuit for-
mation. In this work, we focus on the intrinsic part, which is
not well understood and may occur via several proposed
mechanisms:
1
͑i͒ reversible or physical, e.g., the diffusion of
mobile ions
2
and reorientation of molecular dipoles,
3
and ͑ii͒
irreversible or chemical, e.g., the reactivity of radical ions
4
and reactions of organic materials with oxygen and water
5
͑although only rarely supported by the detection and identi-
fication of degradation products from faded devices͒.
5͑b͒,5͑c͒
The proposed mechanisms can operate on different time
scales and are subject to a large set of variables, such as
device architecture and layer composition, location of the
charge recombination and emission zones, material proper-
ties and purity, fabrication conditions, electrical driving
scheme, operating temperature, etc. Here we focus on the
irreversible intrinsic fade of small-molecule fluorescent
OLEDs.
The chemical instability of the radical cation of the com-
mon light-emitting-layer ͑LEL͒ host aluminum 8-hydroxy-
quinoline chelate ͑Alq
3
͒ has been reported to give rise to
fluorescence quenchers and charge traps that act as nonradi-
ative recombination centers.
6,7
͑Chemical structures are
shown in Fig. 1.͒ Consistent with this proposal, the lifetime
can be increased by mixing Alq
3
with an easier-to-oxidize
material such as an aromatic amine, e.g., N,N
Ј
-bis͑1-
naphthyl͒-N,N
Ј
-diphenylbenzidine ͑NPB͒.
4
This mecha-
nism, however, is unlikely to apply to common blue-emitting
hosts of the anthracene family, e.g., 9,10-bis͑2-naphthyl͒-
2-t-butylanthracene ͑TBADN͒, which form stable radical
ions but nevertheless undergo EL fade and form nonradiative
recombination centers
7
approximately two times faster than
Alq
3
in our hands.
In a simple undoped or doped OLED device, EL report-
edly emanates from a ϳ50–100 Å zone on the LEL side of
the NPB͉LEL interface.
8,9
The charge recombination zone
may be narrower ͑perhaps even a monolayer, i.e., ϳ10 Å͒
because the emission zone is probably enlarged by exciton
diffusion. The EL fade rate is usually roughly proportional to
the drive current density, suggesting that the rate-limiting
step involves either an excited state or a radical ion of an
OLED material. Hence, device lifetime depends on the spe-
cies that carry out the essential LEL functions and the loca-
tion of the charge recombination and emission zones. The
essential LEL functions involve charge transport and recom-
bination, exciton generation and transport, and light emis-
sion. Thus, the lifetime for the most common LEL composi-
tion, a single dopant in a single host, can be increased by the
addition of a certain third LEL component, such as, e.g.,
NPB,
4
TBADN,
10
or rubrene
10,11
for an Alq
3
-based LEL. In
the present work, we ͑i͒ describe a group of PAHs that form
a͒
Electronic mail: viktor.jarikov@kodak.com
JOURNAL OF APPLIED PHYSICS 100, 014901 ͑2006͒
0021-8979/2006/100͑1͒/014901/7/$23.00 © 2006 American Institute of Physics100, 014901-1